exploring the limits of direct mental influence: two ... to prevent a remote mental influence by a...

21
Exploring the Limits of Direct Mental Influence: Two Studies Comparing “Blocking” and “Co-operating” Strategies 1 CAROLINE WATT, JOHN RAVENSCROFT AND ZACHARY MCDERMOTT Koestler Parapsychology Unit, Department of Psychology University of Edinburgh, 7 George Square, Edinburgh EH8 9JZ, UK Abstract — Two studies are reported, with the aim to explore the limits of re- mote mental influence 2 effects. An “influencer” attempts to affect the elec- trodermal activity (EDA) of a distant “influencee”. The influencee is asked to adopt two contrasting mental strategies: for one half of the session, they co- operate with the influence attempt; for the other half they block the influence attempt. Study one, with 32 participants and three experimenter/influencers, found no difference in remote influence effect on EDA between co-operating and blocking conditions (t = - .202), and no overall evidence of a remote mental influence effect (t = - .031). Study two, with 50 participants and two experimenter/influencers, also found no difference for remote influence be- tween co-operating and blocking conditions (t = - .595), but evidence was found of an overall remote influence effect (t = 1.806). Combining the data of the two studies gave an effect size of r = +.13 ( t = 1.176). We also report on participant strategies, an exploratory variance measure of remote influence, sex 3 effects, experimenter effects, and influencer/influencee sex pairings. It is concluded that we have found no evidence that influencee’s mental strategy affects remote mental influence attempts. Thus there is as yet no indication as to the limiting conditions of direct mental influence on living systems. Keywords: parapsychology — psychophysiology — mental strategies — remote influence Introduction In three surveys of concerns over strongly functioning psi, Tart found that pro- fessional parapsychologists (Tart, 1979), Californian students and townspeo- ple (Tart & Labore, 1986), and trainee psychics (Tart, 1986), all expressed fears about possible harmful effects of psi. With a few exceptions, most para- psychologists have given a low profile to considerations of possible harmful effects of psi. Perhaps this is because the sometimes small and inconsistent ef- fect sizes we see in laboratory studies seem quite innocuous. In a recent review Journal of Scientific Exploration, Vol. 13, No. 3, pp. 515–535, 1999 0892-3310/99 © 1999 Society for Scientific Exploration 515 515 1 The first study was previously reported at the 1997 Annual Convention of the Parapsychological Association (Watt et al. , 1997). 2 We use the term ‘influence’ without adopting a particular explanatory framework. Rather, the termi- nology is chosen to reflect how the task is presented to the participant. 3 Although it is considered politically correct to use the term “gender” in preference to “sex,” the for- mer is only appropriate where participants’ gender was actually measured. Our participants were asked to indicate their sex from two categories — male or female — therefore this is a measure of sex not gen- der.

Upload: trinhlien

Post on 22-Jun-2018

216 views

Category:

Documents


0 download

TRANSCRIPT

Exploring the Limits of Direct Mental Influence Two StudiesComparing ldquoBlockingrdquo and ldquoCo-operatingrdquo Strategies1

CAROLINE WATT JOHN RAVENSCROFT AND ZACHARY MCDERMOTT

Koestler Parapsychology Unit Department of PsychologyUniversity of Edinburgh 7 George Square Edinburgh EH8 9JZ UK

Abstract mdash Two studies are reported with the aim to explore the limits of re-mote mental influence2 effects An ldquoinfluencerrdquo attempts to affect the elec-trodermal activity (EDA) of a distant ldquoinfluenceerdquo The influencee is asked toadopt two contrasting mental strategies for one half of the session they co-operate with the influence attempt for the other half they block the influenceattempt Study one with 32 participants and three experimenterinfluencersfound no difference in remote influence effect on EDA between co-operatingand blocking conditions (t = - 202) and no overall evidence of a remotemental influence effect (t = - 031) Study two with 50 participants and twoexperimenterinfluencers also found no difference for remote influence be-tween co-operating and blocking conditions (t = - 595) but evidence wasfound of an overall remote influence effect (t = 1806) Combining the data ofthe two studies gave an effect size of r = +13 (t = 1176) We also report onparticipant strategies an exploratory variance measure of remote influencesex3 effects experimenter effects and influencerinfluencee sex pairings It isconcluded that we have found no evidence that influenceersquos mental strategyaffects remote mental influence attempts Thus there is as yet no indication asto the limiting conditions of direct mental influence on living systems

Keywords parapsychology mdash psychophysiology mdash mental strategies mdashremote influence

Introduction

In three surveys of concerns over strongly functioning psi Tart found that pro-fessional parapsychologists (Tart 1979) Californian students and townspeo-ple (Tart amp Labore 1986) and trainee psychics (Tart 1986) all expressedfears about possible harmful effects of psi With a few exceptions most para-psychologists have given a low profile to considerations of possible harmfuleffects of psi Perhaps this is because the sometimes small and inconsistent ef-fect sizes we see in laboratory studies seem quite innocuous In a recent review

Journal of Scientific Exploration Vol 13 No 3 pp 515ndash535 1999 0892-331099copy 1999 Society for Scientific Exploration

515515

1The first study was previously reported at the 1997 Annual Convention of the ParapsychologicalAssociation (Watt et al 1997)

2We use the term lsquoinfluencersquo without adopting a particular explanatory framework Rather the termi-nology is chosen to reflect how the task is presented to the participant

3Although it is considered politically correct to use the term ldquogenderrdquo in preference to ldquosexrdquo the for-mer is only appropriate where participantsrsquo gender was actually measured Our participants were askedto indicate their sex from two categories mdash male or female mdash therefore this is a measure of sex not gen-der

516 C Watt et al

and meta-analysis of 30 healing analog experiments in which individuals at-tempted to influence autonomic nervous system activity in a distant personhowever the average effect size was r = +25 (Schlitz amp Braud 1997) This ef-fect is not trivial in fact it is approaching a moderate magnitude (Cohen1977) that compares favorably with findings from the more orthodox behav-ioral sciences

Most of these remote influence studies have been framed in a positive orneutral light mdash for example Schlitz and Braudrsquos (1997) review paper is enti-tled ldquoDistant Intentionality and Healingrdquo (italics added) However if experi-mental evidence is provided to show that one person can remotely influencethe physiology of another in a positive or neutral way there is no a priori rea-son why a negative or harmful influence might not also be possible That is itis a logical implication of studies of direct mental influence with living sys-tems that any remote influence effect could potentially be used in a way bene-ficial or harmful to the organism Certainly anecdotal accounts are availableto suggest malignant applications of distant intentionality in real life settings(eg Dossey 1994 Halifax-Grof 1974) and laboratory evidence has been re-ported of distant intentional retardation of bacterial growth (Nash 1982)

It is incumbent on parapsychologists to give more systematic considerationto the question of possible harmful psi for at least two reasons First membersof the public often approach parapsychology laboratories with concerns thatthey may be distantly influenced against their will by another person Thesepeople want advice and information However insufficient research has beendone on this question to enable parapsychologists to give an informed re-sponse Second if the body of evidence supporting the hypothesis thatthoughts are potent continues to mount it is only a matter of time before thelogical implications of this (that thoughts could have harmful as well as bene-ficial effects) trickle through to the media and public consciousness When thishappens parapsychologists will need to have reliable information available torespond to possible concerns

Due to ethical considerations it is difficult to examine the question of harm-ful psi in a realistic way when conducting laboratory research with volunteerparticipants We may approach this question from a different angle howeverto ask whether a person may be able to block or resist an unwelcome mental in-fluence attempt The emphasis here is not on demonstrating the possibility ofharmful psi but on discovering the limiting conditions if any of remote influ-ence effects This brings us to a third reason for conducting such research The-oretically if psi is to be anything more than a vaguely-defined concept thenwe must know something of its limiting conditions (Schlitz amp LaBerge 1997Braud 1985)

Lore exists suggesting how to practice psychic self defense (eg Hope1983) but this question has received very little systematic attention Perhapsmindful of the implications of their successful remote influence studies Braudand Schlitz conducted a preliminary study in which sixteen volunteers at-

tempted to prevent a remote mental influence by a distant observer upon theirown electrodermal activity A second independent group of sixteen cooperat-ed with the remote influence attempt The influencers were to attempt to in-crease the EDA of the target person They were situated in separate rooms re-mote from the target person but received real-time feedback about the targetpersonrsquos EDA through a visual display The influencers were of course un-aware whether the target person was blocking or co-operating In the blockingcondition participants used imagery of various psi-proof screens shields andbarriers and intended that no remote mental influence would occur Overallthere was no evidence of successful blocking of remote mental influencethough significant blocking was found post hoc for one of the two influencers(Braud 1985 Braud et al 1985) A number of studies have looked at concep-tually related questions of psi-mediated helping (Dalton 1994) and attemptedfacilitation or inhibition of performance at forced-choice ESP tasks (Schmei-dler 1958 1961) and clairvoyance tasks (Braud 1985) However there seemsto have been no follow-up to research on direct remote physiological influ-ences

Note that in such physiological blocking studies the influence attempt isdefined as unwelcome only in terms of the volition of the potential influenceethe influence attempt is not itself intended to be harmful but it is the intentionof the influencee to resist such influence nevertheless This paradigm enablesan indirect approach to the question of harmful psi and it also allows us to ex-plore more theoretical questions as to the limits of remote mental influenceWe conducted two studies that were conceptual replications and extensions ofBraud and Schlitzrsquos psi-blocking study The two studies had very similarmethodologies to one another so a detailed description of the method forStudy 1 will be given and we will highlight the differences between the studieswhen we introduce Study 2

Study One Method

The general procedure followed the standard direct mental influence withliving systems (DMILS) method as outlined by Braud and Schlitz (1991) Theinfluencee (participant) was sensorially isolated from the influencer (experi-menter) The influencer monitored the influenceersquos EDA in real time via acomputer display meanwhile following a random influence schedule given bythe computer The schedule told the influencer when to activate calm or haveno influence (rest) on the influenceersquos physiology The influencee was blind tothe instructions for the influence schedule The influencer was blind to the in-fluenceersquos instructions We also administered a vividness of mental imageryquestionnaire and recorded qualitative information on participantsrsquo mentalstrategies to gain some insight into the psychological aspects of resisting a re-mote influence attempt

Blocking vs Co-operating 517

518 C Watt et al

Design

We used a within-subjects design (whereas Braud and Schlitz used a between-subjects design with different groups of volunteers blocking or co-operating)The independent variable was the influenceersquos strategy (blocking vs co-oper-ating) while the influencer attempted either to calm activate or have no influ-ence over the influenceersquos physiology The dependent variable was the influ-enceersquos physiology EDA being the principal measure but skin temperatureand finger pulse volume being recorded for exploratory purposes (we do notreport these exploratory measures in this paper) Our primary aim was to deter-mine whether there was a smaller remote influence effect when the influenceewas blocking compared to when they were co-operating with the influence at-tempt Our secondary aim was to ascertain whether there was any evidence ofan overall remote influence effect

Influence Schedule

The influence schedule that was followed by the influencer was composedof blocks of four epochs mdash either Rest-Activate-Rest-Calm (RARC) or Rest-Calm-Rest-Activate (RCRA) The ordering of each of these blocks of four wasrandomly determined by the computer as the session proceeded so that in totalthere was an equal number of activate and calm epochs Note that the randomsequence also serves to counter-balance the activatecalm conditions acrossthe duration of the study Each epoch was 20 seconds in duration The experi-mental session was thus split into two runs of equal length each having 33recording epochs ie eight blocks of four (RARC or RCRA) plus an extrarest at the end of each run There was a total of 66 recording epochs over theentire session Between the two runs there was a 150 second rest period duringwhich the influencee and the influencer remained in their rooms

The duration of each epoch and the total number of epochs were chosen inorder to maximize the trade-off between gathering sufficient data and keepingthe duration of the session (2 runs of 10 min) at a comfortable length for par-ticipants Additionally we felt that 20 second epochs would be sufficient toenable any stimulus-response effect to manifest in EDA A liberal strategy(that used by Radin et al 1995) was adopted to specify the criteria for includ-ing data if a session was aborted due to technical faults with the psychophysi-ology apparatus data was included from any session that had at least one calmand one activate period (ie a total of four recording epochs)

Data Treatment

Braudrsquos Percentage Score Index (PSI as described in Braud amp Schlitz1991) per session was the unit of analysis used in this study PSI is a conserva-tive summary of the results of an entire session and is calculated by dividingthe sum of EDA activity over all Activate epochs by the sum of all calm andactivate EDA [PSI = S A( S C+ S A)] The rest epochs were not used in this

analysis Their function was primarily to separate the calm and activateepochs from one another and to separate any lingering influence that may havebeen left at the end of the influencerrsquos calm or activate epochs In the absenceof any remote influence effect PSI equals 50 A PSI of greater than 50 rep-resents a remote influence effect in the predicted direction (ie calming whenthat was the influencerrsquos intention activating when that was the influencerrsquosintention) A PSI of less than 50 represents a remote influence effect in thedirection opposite to that predicted For the present study we calculated PSIseparately for each participantrsquos blocking and co-operating condition plus wecalculated an overall PSI score for each participant

Participants

The influencees were volunteer participants drawn from the authorsrsquo con-tacts in the Department of Psychology at Edinburgh University and includingundergraduate psychology students recruited by e-mail and by word of mouthThere were three experimentersinfluencers overall (Watt Ravenscroft andMcDermott) but in each testing session there were only two individuals mdash thevolunteer influencee and the influencer Each experimenter also took part asinfluencer in one session

Written Materials

Each volunteer was given a one page information sheet Once they had readthis they were given a 34-item participant information form (PIF) and a 16-item Vividness of Visual Imagery Questionnaire (Marks 1973) After dis-cussing the procedure with the influencer each participant was asked to sign aconsent form that assured confidentiality and noted their right to withdrawfrom the study at any time Following completion of the DMILS procedure in-fluencee and influencer both completed one-page qualitative post-sessionquestionnaires in which they were invited to describe the kinds of strategiesthat they used during the testing session and to give their subjective impres-sions of the session A letter of thanks and a brief overview of the results weregiven to participants at the end of the study

Technical Apparatus

Two rooms were used for the testing sessions The influenceersquos room wassound-attenuated with double-doors and was located approximately 20 metersaway from the influencerrsquos room down a short flight of stairs The DMILS ap-paratus (an I-410 multichannel physiological monitoring unit from J amp J Sys-tems with source code from Physiodata) was set up between these two roomsand was run on two time-synchronized Pentium computers In the influenceersquosroom a monitor displayed written instructions and a mildly engaging visualdisplay (a screen saver shown during the actual data collection epochs) to theinfluencee This room also contained speakers that gave auditory instructions

Blocking vs Co-operating 519

520 C Watt et al

(ldquohey block nowrdquo ldquohey co-operate nowrdquo and ldquosession over thank yourdquo) tothe influencee just prior to each of the two 10-minute runs duplicating thewritten instructions that appeared on the computer monitor The physiologysensors that were attached to the influenceersquos fingers were located in the influ-enceersquos room The signals from the sensors were relayed along a cable andprocessed by a computer located in the influencerrsquos room Therefore the com-puter in the influenceersquos room was merely used to provide instructions and avisual display to the influencee The computer in the influencerrsquos room con-trolled the DMILS software and processes and stored the psychophysiologydata

Procedure

The influencer and participant initially sat down in the reception room in theparapsychology suite for a further description of procedure and returned theircompleted PIFimagery questionnaire to the influenceer who went over it withthe participant Any questions that were raised by the participant prior to test-ing were answered here (This question-and-answer session usually lasted fif-teen minutes) The participant then gave their informed consent to the studyThe influencer also suggested various mental approaches the participant mightadopt for the blocking and co-operating conditions (some examples weregiven in the information sheet) These suggestions included both visual im-agery (eg visualizing a shielding cocoon in the blocking condition) and cog-nitive strategies (eg adopting a stubborn and uncooperative frame of mind inthe blocking condition) Participants were informed that different strategiesmight suit different people and they were to do whatever felt natural for themThe influencer then took the participant to the testing room and demonstratedthe auditory instructions so that the participant would not be startled by thesewhen they occurred during the session itself The participant was also taken tosee where the influencer was sitting and the influencer described what theywould be doing when they were attempting to activate or calm the participant

Once the preliminaries were finished the influencee was seated in a partiallyreclined chair in front of the computer monitor and the physiology sensorswere attached to their fingers by the influencer The room was softly illuminat-ed by a desk lamp with an orangered filter At this stage the influencer checkedthat the physiology signals were being transmitted to the computer in theirroom Returning to the influencee the influencer set a start time on the influ-enceersquos computer This was on average two to three minutes later The influ-encer then returned to their room closing all doors and waited for the start ofthe session

Physiology Data Collection As the session began the influencee was in-structed visually and auditorially either to ldquohey block nowrdquo or ldquohey co-oper-ate nowrdquo The influenceersquos monitor then displayed a colorful and dynamicscreen-saver that was intended to be engaging without being particularlyarousing or relaxing At the same time the influencer was following the influ-

ence schedule as instructed by the computer observing the real-time feedbackdisplay of the participantrsquos EDA when desired At the end of the first run whichlasted approximately ten minutes the influencer saw a 150 second count downon the computer before the second run began At the same time the influenceesaw a message about the strategy to adopt for the forthcoming second run (ieto block or to co-operate whichever strategy was not used for the first run) Atthe beginning of the second run the influencee was also instructed auditoriallyeither to ldquohey block nowrdquo or ldquohey co-operate nowrdquo The dynamic visual dis-play resumed while the influencer followed a random influence schedule forapproximately ten minutes The entire session took about 25 minutes to com-plete At the end of the session the influencee was auditorially and visually in-formed ldquosession over thank yourdquo and waited until the influencer returned tothe influenceersquos room

Post-Testing Procedures The influenceersquos physiology sensors were re-moved by the influencer and both returned to the reception room to completethe post-session questionnaires Then the influencee was taken to the influ-encerrsquos room to see feedback on the results of their testing session There wasan option for some final discussion about the session after which the volun-teer was thanked for their participation and left Once all the data was collect-ed feedback on the results of the study as a whole was sent to participants

Planned Hypotheses and Analyses

Our main aim was to examine whether there was any evidence for blockingof a remote mental influence on physiology [EDA is the principal physiologymeasure here with per-session PSI scores representing the measure of theDMILS effect on EDA] H1 therefore predicted significantly greater PSIscores in the co-operating condition than in the blocking condition (using a re-lated t-test one-tailed) Our second question was whether there was any over-all evidence of remote mental influence on EDA This analysis would be con-ducted on the data of blocking and co-operating runs combined if nosignificant difference was found between these two conditions (as tested byH1) If H1 was supported H2 would be tested on the data of co-operating runsonly predicting PSI scores significantly greater than chance expectation(using a single-mean t-test of the PSI score vs MCE one-tailed)

Planned alpha level was 05 There was no correction for multiple analysisThis was dealt with by restricting the number of formal hypotheses (with one-tailed significance tests) to two in each study All other analyses were regardedas exploratory and two-tailed significance tests were used

Exploratory Measures

Variance

DMILS researchers do not yet appear to have identified what might be themost appropriate indicator of a remote influence effect Therefore in addition

Blocking vs Co-operating 521

522 C Watt et al

to the PSI score we looked at the amount of variance that was shown in block-ing and co-operating scores The variance measure was calculated as followsMean values for activate periods and calm periods were calculated separatelyfor block and co-operate conditions Mean values were taken from each indi-vidual score in that condition and squared and summed S [mean value for con-dition - individual score]2 This gave a summation of variability around themean for each activate calm and rest period for block and cooperate condi-tions It was used in conjunction with the mean level of EDA activity givingan indication of the motion of the wave (ie the variability of the EDA activi-ty) rather than the overall level

VVIQ

Anecdotally the use of appropriate imagery (eg visualize being surroundedby a shielding white halo of white light) is suggested to be an effective blockingtechnique (eg Hope 1983) We therefore wished to see if participants who re-ported experiencing vivid imagery with ease are perhaps more effective block-ers than those who do not report clear mental imagery Descriptive data wasgathered to explore any trends in PSI scores for blockingco-operating forthose participants with VVIQ scores ranging from low (ie reporting vividmental imagery) to high (ie reporting little or no mental imagery)

Qualitative Reports

We also wanted to report data illustrating what kind of different strategiesparticipants used when they were blocking and co-operating These qualitativereports were gathered in the form of the participantsrsquo written responses toopen-ended questions on a post-session questionnaire Watt and Ravenscroftindependently coded these written responses into categories and cross-checked to ensure coding agreement

Sex and Experimenter Effects

We also examined the data for sex effects and experimenter effects becausethese variables could confound the interpretation of any overall effects Final-ly a post hoc analysis was conducted on influenceeinfluencer sex pairings Ina study on the unconscious detection of remote staring Schlitz and LaBerge(1997) found the greatest effect for opposite-sex observer-subject pairs

Study One Results and Discussion

Participants

Thirty-two individuals took part as influencees (12 females 20 males) theirages ranged from 19 to 63 years (mean age 285 years)

Incomplete Data

One participant declined to give their age on the PIF form so the descriptivestatistics on age do not include this participant (so N = 31) Also due to a tech-nical hitch the exploratory variance scores were lost for one participant so N = 31 for this measure There were no drop-outs from this study

Remote Influence Measure

Table 1 presents a summary of the remote influence data in the form ofmean Percentage Score Index for co-operating and blocking conditions Therewas no significant difference between PSI for co-operating and blocking con-ditions (t = - 202 df = 31) Therefore Hypothesis 1 (predicting significantlygreater PSI scores for co-operating condition compared to blocking condition)was not supported by these results As planned the PSI scores for co-operatingand blocking runs were combined to examine Hypothesis 2 (of an overall re-mote mental influence effect in line with the calming and activating intentionsof the influencee) Mean PSI scores were exactly at chance expectation (t = - 031 df = 31) therefore there was no support for Hypothesis 2

Exploratory Questions

Variance Measure

Table 2 presents summary data for the exploratory measure of the amount ofvariance shown in each condition (lt 50 represents reduced variance gt 50 rep-resents increased variance) There was no significant difference between vari-ance scores for co-operating and blocking conditions (t = 835 df = 30) andwhen the conditions were combined variance did not depart from chance ex-pectation (t = - 244 df = 30) It is interesting to note however that the vari-ance measure gives a much wider range of scores than the PSI measure

Blocking vs Co-operating 523

TABLE 1Summary Data Mean PSI Scores

(N = 32 MCE = 50 range in brackets)

Co-operate PSI Block PSI Combined PSI

4996 5003 5000(4692- 5215) (4653- 5404) (4853- 5309)

TABLE 2Summary Data Mean Variance Scores(N = 31 MCE = 50 range in brackets)

Co-operate Variance Block Variance Combined Variance

5089 4864 4977(2648- 7012) (2492- 6484) (4116- 6297)

524 C Watt et al

suggesting that the variance measure may potentially be a more sensitive indi-cator of remote influence effect than PSI

Vividness of Mental Imagery

Low VVIQ scores represent participants reporting highly vivid imagery andhigh VVIQ scores represent those reporting little or no imagery The meanscore was 4275 (range from 21 to 80) Using a Pearson correlation we foundthat there was no evidence to suggest that those participants with low VVIQscores (high imagery) were better blockers than those with high scores (VVIQvs PSI blocking r = - 113) We also found that there was no evidence of a cor-relation between vividness of imagery and overall remote influence effectsusing the combined Percentage Score Index measure (VVIQ vs PSI overall r =- 077)

Influenceesrsquo Blocking and Co-operating Strategies

Influencees completed a questionnaire (after the session was over) askingthem to describe what strategies they used for blocking and co-operating Cat-egorizing these responses we found a difference in the kinds of strategies usedwhen blocking compared to co-operating as shown in Table 3 In the blockingcondition the majority of participants (469) used visual imagery strategiesFewer (281) participants used a combination of imagery and cognitivestrategies (such as repeating words related to blocking) while 25 reportedusing purely cognitive strategies for blocking (such as adopting a resistant anduncooperative attitude or thinking of irrelevancies) In the co-operating con-dition fewer participants used visual imagery alone (333) The majority ofparticipants (424) reported that their strategies for co-operating includedadopting a feeling of passiveness openness or relaxation whereas no partici-pants adopted this kind of strategy for the blocking condition 121 of partic-ipants used cognitive strategies alone (such as adopting a co-operative atti-tude) and the same proportion used a combination of visual imagery andcognitive strategies

Therefore there was a tendency for participants to use visual imagery forblocking and relaxationpassivity or visual imagery for co-operating This wasdespite the fact that we asked participants to attempt to adopt similar levels ofactivation for both co-operating and blocking strategies There was also a ten-dency not surprisingly for those participants with low VVIQ scores (ie vivid

TABLE 3Influenceesrsquo Mental Strategies ()

Vis Im Cognitive VisCog OpenRelax

Blocking 469 250 281 00Co-operating 333 121 121 424

imagers) to use visual imagery strategies Those who had high VVIQ scoresalso used visual imagery but more often used this in combination with othertactics

Other Qualitative Reports from Participants According to their post-ses-sion questionnaire responses participants generally felt comfortable about theattempt to be influenced (N = 24) Three participants said they felt apprehen-sive or uncomfortable with the idea of being influenced Interestingly of thesethree two were actively involved with parapsychology and were aware of theharmful psi implications discussed at the beginning of this report The thirdwas a psychology student who had attended a lecture given by Watt in whichthe possibilities of harmful psi were discussed Another group of three partici-pants said they would be uncomfortable with the idea of being influenced ifthey believed such influence was possible

Seven participants said they found it was difficult to sustain concentrationfor the 10 minutes that comprised each condition Five participants found thescreensaver distracting or disturbing Two participants specifically mentionedthat they found blocking a more taxing task than co-operating

Sex Effects

There were more males (N = 20) than females (N = 12) in this study howev-er there was no significant difference in PSI scores for remote physiology in-fluence between males and females (male PSI vs female PSI t = 1085 male PSI = 4987 female PSI = 5021)

Experimenter Effects

We studied whether there was any difference in PSI scores obtained by thethree experimenters The results are shown in Table 4 There seems to be nosign of experimenter effects in study one

InfluenceeInfluencer Sex Pairings

Table 5 gives the mean PSI scores for different influenceeinfluencer pair-ings Taking the combined PSI scores it can be seen that the highest remote in-fluence effect is for FM pairs followed by MF pairs followed by FF pairsand the lowest scoring is for MM pairs Post hoc it was found that the mixedsex pairings showed significantly higher remote influence scores than the same

Blocking vs Co-operating 525

TABLE 4Remote Influence Results by Experimenter

Experimenter N sessions Coop PSI Block PSI Combined PSI

CW 12 4983 5011 4997ZMcD 10 5015 4998 5006JR 10 4994 4999 4996

526 C Watt et al

sex pairings (t = - 2016 df = 30 p = 053 2 - t) This replicates the effect foundby Schlitz and LaBerge (1997)

Study One Conclusions

In Study 1 we developed a methodology to examine intentional resistance toa remote mental influence attempt (Watt et al 1997) We contrasted influ-enceesrsquo physiological activity when they were mentally co-operating with aremote attempt to influence their physiology with physiological activity wheninfluencees were mentally blocking the influence attempt We found no evi-dence for any remote influence effect overall and no difference betweenblocking and co-operating conditions Our qualitative measures yielded someinteresting information about participantsrsquo mental strategies and reactions dur-ing the remote influence session We found no difference between experi-menters in remote influence scores but we did find that mixed sex influ-enceeinfluencer pairings scored more highly than same sex pairings

Study Two Method

Study 2 was planned to explore in more depth the nature of the relationshipbetween aspects of personality and cognition and psi performance Partici-pants took part in three testing sessions approximately one week apart Thefirst consisted of various psychological measures the second consisted of aforced choice ESP task and the third consisted of a replication of the blocking-co-operating study (Study 1) described above The psychological measureswere not scored until after completion of the remote influence session Partici-pants received feedback about their forced choice ESP performance at the con-clusion of the second session For ease of interpretation the present paper isrestricted to a description of the third testing session which was essentially areplication of study one with fewer influencers and greater statistical power

We introduced some slight changes to the methodology of Study 2 based onour experiences with Study 1 At the start of each of the two runs in Study 1the participant was given auditory instructions either to ldquoblock nowrdquo or ldquoco-operate nowrdquo We observed that the auditory instructions appeared to affect theparticipantrsquos EDA as if they were startled As planned the first recordingepoch in each run was not used in analysis to give the participantrsquos EDA time

TABLE 5Mean PSI Scores for Different InfluenceeInfluencer Sex Pairings

Pairing N Coop PSI Block PSI Combined PSI

M-F 6 4932 5069 5001F-M 6 5043 5034 5048F-F 6 5022 4952 4993M-M 14 4988 4975 4981

to settle down However in order to remove any remaining potential artifactu-al effects on EDA in Study 2 only visual instructions were given

Another difference in methodology concerned the screensaver viewed bythe participants during the experimental session In Study 1 some participantsreported that the series of patterns produced by the computer were being dis-played at too fast a rate such that some participants found it distracting or dis-turbing Again we felt that this may have an artifactual effect on EDA so wedecided for study two to reduce the speed of the screensaver

Thirdly in Study 2 we did not give participants detailed information abouttheir remote influence performance until the feedback letters that were distrib-uted at the conclusion of the entire study The reason for this change was mere-ly to streamline the procedure on what was already a time-consuming experi-ment Finally in another streamlining change we did not take VVIQ measuresin Study 2

Planned Hypotheses and Analyses

These were the same as with Study 1 That is hypothesis 1 predicted higherPSI scores for the co-operating condition than for the blocking condition(measured with a related t-test one-tailed) If no significant difference wasfound between blocking PSI and co-operating PSI scores the two would becombined to look at overall remote influence effect (hypothesis 2 single-meant-test one-tailed) If a significant difference was found with hypothesis 1 theanalysis for hypothesis 2 would be conducted with the data for co-operatingsessions only

Exploratory Questions

As with study one we took an exploratory variance measure of remote EDAinfluence We also gathered qualitative descriptions of the mental strategiesparticipants adopted whilst blocking and co-operating We also looked at sexeffects experimenter effects and post hoc influenceeinfluencer sex pairings

Participants

Study 2 had two experimenterinfluencers (Watt amp Ravenscroft) After aparticipant was recruited (usually by Ravenscroft) they were assigned to oneof the two experimenters for all three testing sessions This assignment wasdone unsystematically according to the preferences and availability of the ex-perimenters As planned 50 volunteer participants were drawn from the gener-al participant pool at the Koestler Parapsychology Unit and by word ofmouth All participants completed the Participant Information Form and aConsent form prior to the study The participants consisted of 33 females and17 males Average age was 41 years range from 22 years to 76 years of age Inorder to increase the studyrsquos relevance to the general population undergradu-ate psychology students were not recruited as participants The main criterion

Blocking vs Co-operating 527

528 C Watt et al

for participation was that the participants were interested in and open-mindedabout parapsychology Again there were no drop-outs from this study Allparticipants received a detailed personal feedback letter at the conclusion ofthe study

Study Two Results and Discussion

Remote Influence Measure

PSI scores are summarized in Table 6 which shows that the mean PSI scorewhen co-operating was nonsignificantly lower than mean PSI score whenblocking (t = - 595 df = 49) Hypothesis 1 was therefore not supported Asplanned in the absence of any significant difference between blocking and co-operating PSI scores the two conditions were combined to test hypothesis 2which predicted overall significant remote mental influence performance Thecombined PSI score was significantly greater than mean chance expectation of50 (t = 1806 df = 49 p = 038 1 - t) Therefore hypothesis 2 was supportedgiving evidence of an overall remote influence effect upon psychophysiologyin the prespecified direction

The failure to support hypothesis 1 confirms what we found in Study 1 thatthere was no significant difference between blocking and co-operating condi-tions on remote influence scores Indeed both studies found a slight effect inthe opposite direction to that predicted that is slightly higher PSI scores whenblocking was the influenceersquos intention than when co-operating was their in-tention It was encouraging to find significant support for a remote influenceeffect in the predicted direction for blocking and co-operating conditions com-bined No such trend was found in the previous study though with a smallernumber of participants (N = 32) the earlier study had lower statistical powerthan the present study

Exploratory Questions

Variance Measure

Table 7 shows the results for the exploratory measure of EDA variance lt 50represents reduced variance gt 50 represents increased variance We found nodifference between variance scores for cooperate versus block conditions(t = 150 df = 49) When these are combined the resulting figure does not de-viate from chance expectation (t = 440 df = 49) Thus as in Study 1 it seems

TABLE 6Summary Data Mean PSI Scores

(N = 50 MCE = 50 range in brackets)

Co-operate PSI Block PSI Combined PSI

5008 5019 5016(4821- 5221) (4807- 5557) (4900- 5296)

that the exploratory variance measure does not reveal any additional indica-tion of the operation of a remote mental influence in this study However thewider range of variance scoring continues to suggest that this measure mayhave greater potential sensitivity than the summary PSI score

Influenceesrsquo Blocking and Co-operating Strategies

We were interested in learning more about what particular strategies partici-pants chose to adopt when blocking and co-operating As in Study 1 a post-session questionnaire was used to gather this information Table 8 shows thepercentage of participants who adopted each kind of strategy

The table shows a very similar pattern to that found in Study 1 The mainpoint of interest is that almost no participants took a passive approach toblocking (ie adopting an openrelaxing strategy) A much higher proportionpreferred to adopt this approach when co-operating The preferred mode forboth conditions was to use some form of visual imagery such as imagining ashield or a screen when blocking and imagining an open channel such as ariver when co-operating

The majority of participants reported no concerns over the possibility ofhaving their physiology remotely influenced by another person as found instudy one

Sex Effects

As in study one the sexes were not equally represented among participantsWhereas Study 1 had an excess of male participants in Study 2 twice as manyfemales (N = 33) as males (N = 17) participated It was therefore important tolook for sex differences since any such differences could confound the inter-pretation of overall trends We found no difference between sexes when co-op-erating (female mean PSI score = 5008 male mean PSI score = 5006 t = 107 df = 48) There was a nonsignificant trend for females to have higher

Blocking vs Co-operating 529

TABLE 7Summary Data Mean Variance Scores(N = 50 MCE = 50 range in brackets)

Co-operate Variance Block Variance Combined Variance

5076 5054 5034(3164- 7279) (3572- 6813) (3749- 6671)

TABLE 8Influenceesrsquo Mental Strategies ()

Visual Imagery Cognitive VisCog OpenRelax

Blocking 64 26 6 2Co-operating 60 8 12 20

530 C Watt et al

PSI scores than males when blocking (mean PSI scores 5030 and 4998 re-spectively t = 992 df = 48) Combining blocking and co-operating conditionsthere was no overall difference between the sexes (mean PSI scores 5020 forfemales 5008 for males t = 665 df = 48)

Experimenter Effects

Study 2 found a statistically significant overall remote influence effect(combined PSI = 50155 t = 1806) Table 9 shows the breakdown of remoteinfluence scores for the two experimenters CW and JR Looking at the overalldata for the two experimenters we see that CWrsquos sessions were associatedwith most of the remote influence effect However due to the smaller N CWrsquossessions are not independently significant (t = 1632) and there is not a signifi-cant difference between CW and JRrsquos overall PSI scores (t = 678) Post hocwe calculated effect size r according to the formula r = [t2(t2+df)]12 for the twoexperimenters For JR (t = 817 df = 22) we found a small effect sizer = +17 For CW (t = 1632 df = 26) the effect size was moderate r = +30This may be some post hoc indication of an experimenter effect in Study 2

InfluenceeInfluencer Sex Pairings

As can be seen from Table 10 Study 2 found exactly the opposite trend tothat found in Study 1 On a post hoc test same sex pairings scored significant-ly higher on the remote influence PSI measure than mixed sex pairings(t = - 2034 df = 48 p = 048 2 - t) This contradicts Schlitz and LaBergersquos(1997) finding of greater autonomic detection of remote staring for mixed sexpairs

TABLE 9Remote Influence Results by Experimenter

Experimenter N sessions Coop PSI Block PSI Combined PSI

CW 27 5011 5026 5021JR 23 5003 5010 5009

TABLE 10Mean PSI Scores for Different InfluenceeInfluencer Sex Pairings

Pairing N Coop PSI Block PSI Combined PSI

MF 11 4977 5009 4997FM 17 4983 5021 5003FF 16 5035 5038 5038MM 6 5060 4978 5027

General Summary and Conclusions

We have described the findings of two studies exploring the limits of remotemental influence effects We wished to see whether the mental strategiesadopted by the influencee would affect remote influence scores In particularcould an influencee block or resist a remote influence attempt We recorded in-fluenceesrsquo electrodermal activity while the influencer attempted to calm to ac-tivate or to have no influence over the influenceesrsquo physiology In each studythere were two conditions reflecting the influenceesrsquo designated mental strate-gies in response to the remote influence attempt blocking and co-operating

There were only two formal hypotheses 1) that remote influence scoreswould be significantly higher during the co-operating condition than duringthe blocking condition and 2) that overall there would be significant evidenceof remote mental influence on physiology in the designated direction (iecalming when calming was the influencerrsquos intention activating when activat-ing was the influencerrsquos intention) Contrary to prediction Study 1 found thatremote influence effects were higher in the block condition than in the co-op-erate condition (but not to a significant extent) Also overall remote influenceperformance was at chance level This study therefore found no support for ei-ther of the formal hypotheses Study 2 consisted of a replication of Study 1with a few minor adjustments plus additional measures that were not reportedin this paper Again contrary to prediction remote influence scores were non-significantly higher in the block condition than in the co-operate conditionThis replicates the trend found in Study 1 failing to support hypothesis oneThese results suggest that the mental strategy adopted by the influencee in theface of a remote influence attempt has no effect on remote influence perfor-mance Referring back to the questions posed in the introduction then wehave not found any evidence to suggest that it is possible to resist a remote in-fluence effect and there is as yet no indication of the possible limiting condi-tions for remote mental influence

Unlike Study 1 Study 2 found statistically significant evidence of an overallremote influence effect in the prespecified direction (ie calming of EDAwhen that was the influencerrsquos intention activation of EDA when that was theinfluencerrsquos intention) In order to facilitate comparison with a recent meta-an-alytic review of remote influence studies we calculated the effect size (r) ac-cording to the formula r = [t2(t2+df)]12 For their review of 19 EDA influencestudies Schlitz and Braud (1997) found an average effect size of r = +25 OurStudy 1 with overall t = - 031 df = 31 gives a negligible effect size of r = - 006 Study 2 with overall t = 1806 df = 49 gives r = +25 Study 2 there-fore replicates the mean effect size found in EDA studies conducted by otherresearchers in other parapsychology laboratories Post hoc we can combinethe data of the two studies for greater statistical power Overall t = 1176 df = 81 giving a small effect size of r = +13

We can find no clear reason why Study 2 found an EDA influence effectwhile Study 1 did not A number of factors are worth considering 1) The

Blocking vs Co-operating 531

532 C Watt et al

significant result could be due to chance or undetected artifact 2) Study 2 im-plemented what we considered to be minor methodological improvements thatmight have removed some artifactual fluctuations from the EDA data that weobserved in study one 3) Each experimenter in Study 2 conducted three testingsessions with their participant By the third session in which the remote influ-ence measure was taken experimenters were on quite friendly terms with theirparticipants perhaps more so than in Study 1 where each participant came infor only one testing session 4) The participants in Study 2 were more represen-tative of the general population we felt than those in Study 1 Study 1 in-volved more undergraduate psychology students and associates of the experi-menters 5) The experimenters in study one were less experienced with theremote influence protocol than when they conducted Study 2

In addition to the formal hypotheses we reported a number of exploratorymeasures Study 1 found that influenceesrsquo self-reported vividness of visual im-agery appeared to be unrelated to the remote influence performance In bothstudies influenceesrsquo qualitative descriptions of the actual mental strategiesthey employed when blocking and co-operating revealed a tendency for differ-ent strategies to be preferred for the different conditions The use of visual im-agery was popular for both blocking (eg visualizing being surrounded by aprotective cocoon) and co-operating (eg visualizing a river flowing betweeninfluencer and influencee) However a considerable proportion of participantsreported adopting an open or relaxed frame of mind when co-operating whilstalmost no participants chose this strategy when blocking This was despite thefact that prior to each session we encouraged participants to adopt comparablestrategies for blocking and co-operating

It was interesting and perhaps surprising to note that in both studies themajority of participants reported no feelings of concern or discomfort at thepossibility of having their physiology remotely influenced Those who wereconcerned tended to be those who knew more about parapsychology and thepossible harmful applications of remote influence A small proportion of par-ticipants was unconcerned because they did not believe such an influence waspossible in the first place

Both studies employed an exploratory variance measure of EDA influenceThis measure showed a greater range of scoring overall thus perhaps greaterpotential sensitivity to a remote influence effect However the variance mea-sure showed no indication of a remote influence effect To date researchershave employed a variety of indicators of remote influence as the literaturegrows progress may be made on identifying which is the most appropriate in-dicator

Neither study found significant sex differences in remote influence effectsthough in both studies female influencees tended to have higher remote influ-ence scores than male influencees

In terms of their overall remote influence effects there appeared to be nodifference between experimenterinfluencers in Study 1 (r = +06 - 05 and

- 06) In Study 2 effect size measures indicated that the experimenters dif-fered in remote influence effect sizes (r = +17 and +30) Due to the smaller Nthis difference was not found to be statistically significant Post hoc combin-ing Studies 1 and 2 for experimenters JW and CW (ZMcD was not an experi-menter in Study 2) the effect sizes are r = +08 and r = +20 respectively

We conducted a post hoc analysis of the effects of influenceeinfluencer sexpairings on remote influence scores Our two studies came up with contradic-tory findings each statistically significant Study one found that mixed sexpairings had significantly higher remote influence scores than same sex pair-ings This replicated the post hoc finding of Schlitz and LaBerge (1997) in con-nection with the autonomic detection of remote observation Study 2 foundthat same sex pairings had significantly higher remote influence scores thanmixed sex pairings Our contradictory findings do not help to clarify the pic-ture on the importance or otherwise of sex pairings It is quite possible that sexis not the most salient aspect of the dynamic between pairs Age personalityand familiarity are likely to be important variables moderating the relationshipbetween influencer and influencee

There are many more variables that we could look at but as the number ofanalyses multiplies so does the opportunity for spurious significant resultsFor this reason we have named only two formal hypotheses for each experi-ment All other analyses are regarded as exploratory and any trends that wehave highlighted above would require replication

In conclusion these two studies have found no evidence of a difference inremote influence effects when influencees are asked to adopt cooperative men-tal strategies compared to when they are asked to adopt blocking mental strate-gies In fact contrary to expectation both studies found slightly higher remoteinfluence effects for the blocking condition compared to the co-operating con-dition On this basis we have found no indication of limiting conditions for re-mote influence effects Therefore while anecdotal advice exists on how to de-fend oneself against a remote physiological influence attempt in the realworld laboratory evidence to date has provided no support for this advice

Given the importance of this question both in practical terms and in theoret-ical terms we argue that this line of research deserves continued investigationIf a way could be found to conduct research with a more realistic scenario mdash where the remote influence was not just neutrally and abstractlydefined as unwelcome mdash then our research might have more meaningful thingsto say about the parameters of remote influence It is interesting and perhapssurprising to note that when Watt has lectured on this topic to informed lay au-diences the feedback has often been that researchers must use more potential-ly harmful or disruptive remote influence (of course with the informed consentof the participants) While there are clear ethical inhibitions when consideringpossible physiological influence perhaps research using higher level influencewould be more acceptable For example one could attempt to disrupt or en-hance the concentration of chess players who were aware of the influence

Blocking vs Co-operating 533

534 C Watt et al

attempt and who were concurrently blocking or co-operating with the influ-ence attempt Research already exists to suggest that remote influence does notjust operate at physiological levels Attention focusing for example can be fa-cilitated by a remote helper (Brady amp Morris 1997 Braud et al 1995) Futureresearch in this area must therefore expand beyond physiological influences

Acknowledgements

The authors gratefully acknowledge the financial support they receivedfrom the Institut fuumlr Grenzgebiete der Psychologie und PsychohygieneFreiburg Germany while conducting these studies We would also like tothank Dr Paul Stevens for generously helping to solve software difficultieswhen preparing for these studies and the anonymous reviewer for helpfulcomments Address reprint requests to Dr C Watt Department of PsychologyUniversity of Edinburgh 7 George Square Edinburgh EH8 9JZ Scotland UK

ReferencesBrady C amp Morris R L (1997) Attention focusing facilitated through remote mental interac-

tion A replication and exploration of parameters In R Wiseman (Ed) Proceedings of Pre-sented Papers The Parapsychological Association 40th Annual Convention held in conjunc-tion with The Society for Psychical Research pp 73ndash91

Braud W G (1985) Blockingshielding psychic functioning through psychological and psychictechniques A report of three preliminary studies In RA White and J Solfvin (Eds) Re-search in Parapsychology 1984 Metuchen NJ Scarecrow Press pp 42ndash44

Braud W G amp Schlitz M J (1991) Consciousness interactions with remote biological systemsAnomalous intentionality effects Subtle Energies 2 1

Braud W G Schlitz M J Collins J and Klitch H (1985) Further studies of the bio-PK effectFeedback blocking specificitygenerality In RA White amp J Solfvin (Eds) Research inParapsychology 1984 Metuchen NJ Scarecrow Press pp 45ndash48

Braud W G Shafer D McNeill K and Guerra V (1995) Attention focusing facilitatedthrough remote mental interaction Journal of the American Society for Psychical Research89 103

Cohen J (1997) Statistical Power Analysis for the Behavioral Sciences New York AcademicPress

Dalton K (1994) Remotely influenced ESP performance in a computer task A preliminarystudy Proceedings of the 37th Annual Convention of the Parapsychological Association pp95ndash103

Dossey L (1994) Healing and the mind Is there a dark side Journal of Scientific Exploration 873

Halifax-Grof J (1974) Hex death In A Angoff amp D Barth (Eds) Parapsychology and Anthro-pology Proceedings of an International Conference Held in London England August 29ndash311973 NY Parapsychology Foundation

Hope M (1983) Practical Techniques of Psychic Self-Defence Wellingborough Northampton-shire Aquarian Press

Marks DF (1973) Visual imagery differences in the recall of pictures British Journal of Psy-chology 64 17

Nash CB (1982) Psychokinetic control of bacterial growth Journal of the Society for PsychicalResearch 51 217

Radin DI Taylor RK and Braud WG (1995) Remote mental influence of human electroder-mal activity A pilot replication European Journal of Parapsychology 11 19

Schlitz M amp Braud W G (1997) Distant intentionality and healing Assessing the evidence Al-ternative Therapies 3 62

Schlitz M J amp LaBerge S (1997) Covert observation increases skin conductance in subjectsunaware of when they are being observed A replication Journal of Parapsychology 61 185

Schmeidler G R (1958) Agent-percipient relationships Journal of the American Society forPsychical Research 52 47

Schmeidler G R (1961) Evidence for two kinds of telepathy International Journal of Parapsy-chology 3 5

Tart C T (1979) A survey of expert opinion on potentially negative uses of psi United Statesgovernment interest in psi and the level of research funding of the field In WG Roll (Ed) Re-search in Parapsychology 1978 Metuchen NJ Scarecrow Press pp 54

Tart C T amp Labore C M (1986) Attitudes towards strongly functioning psi A preliminary sur-vey Journal of the America Society for Psychical Research 80 163

Tart C T (1986) Psychicsrsquo fears of psychic powers Journal of the American Society for Psychi-cal Research 80 278

Watt C A McDermott Z amp Ravenscroft J (1997) Exploring the limits of remote mental inter-action A comparison of blocking versus co-operating strategies In R Wiseman (Ed) Pro-ceedings of the 40th Annual Convention of the Parapsychological Association held in conjunc-tion with the Society for Psychical Research Brighton England pp 464ndash476

Blocking vs Co-operating 535

516 C Watt et al

and meta-analysis of 30 healing analog experiments in which individuals at-tempted to influence autonomic nervous system activity in a distant personhowever the average effect size was r = +25 (Schlitz amp Braud 1997) This ef-fect is not trivial in fact it is approaching a moderate magnitude (Cohen1977) that compares favorably with findings from the more orthodox behav-ioral sciences

Most of these remote influence studies have been framed in a positive orneutral light mdash for example Schlitz and Braudrsquos (1997) review paper is enti-tled ldquoDistant Intentionality and Healingrdquo (italics added) However if experi-mental evidence is provided to show that one person can remotely influencethe physiology of another in a positive or neutral way there is no a priori rea-son why a negative or harmful influence might not also be possible That is itis a logical implication of studies of direct mental influence with living sys-tems that any remote influence effect could potentially be used in a way bene-ficial or harmful to the organism Certainly anecdotal accounts are availableto suggest malignant applications of distant intentionality in real life settings(eg Dossey 1994 Halifax-Grof 1974) and laboratory evidence has been re-ported of distant intentional retardation of bacterial growth (Nash 1982)

It is incumbent on parapsychologists to give more systematic considerationto the question of possible harmful psi for at least two reasons First membersof the public often approach parapsychology laboratories with concerns thatthey may be distantly influenced against their will by another person Thesepeople want advice and information However insufficient research has beendone on this question to enable parapsychologists to give an informed re-sponse Second if the body of evidence supporting the hypothesis thatthoughts are potent continues to mount it is only a matter of time before thelogical implications of this (that thoughts could have harmful as well as bene-ficial effects) trickle through to the media and public consciousness When thishappens parapsychologists will need to have reliable information available torespond to possible concerns

Due to ethical considerations it is difficult to examine the question of harm-ful psi in a realistic way when conducting laboratory research with volunteerparticipants We may approach this question from a different angle howeverto ask whether a person may be able to block or resist an unwelcome mental in-fluence attempt The emphasis here is not on demonstrating the possibility ofharmful psi but on discovering the limiting conditions if any of remote influ-ence effects This brings us to a third reason for conducting such research The-oretically if psi is to be anything more than a vaguely-defined concept thenwe must know something of its limiting conditions (Schlitz amp LaBerge 1997Braud 1985)

Lore exists suggesting how to practice psychic self defense (eg Hope1983) but this question has received very little systematic attention Perhapsmindful of the implications of their successful remote influence studies Braudand Schlitz conducted a preliminary study in which sixteen volunteers at-

tempted to prevent a remote mental influence by a distant observer upon theirown electrodermal activity A second independent group of sixteen cooperat-ed with the remote influence attempt The influencers were to attempt to in-crease the EDA of the target person They were situated in separate rooms re-mote from the target person but received real-time feedback about the targetpersonrsquos EDA through a visual display The influencers were of course un-aware whether the target person was blocking or co-operating In the blockingcondition participants used imagery of various psi-proof screens shields andbarriers and intended that no remote mental influence would occur Overallthere was no evidence of successful blocking of remote mental influencethough significant blocking was found post hoc for one of the two influencers(Braud 1985 Braud et al 1985) A number of studies have looked at concep-tually related questions of psi-mediated helping (Dalton 1994) and attemptedfacilitation or inhibition of performance at forced-choice ESP tasks (Schmei-dler 1958 1961) and clairvoyance tasks (Braud 1985) However there seemsto have been no follow-up to research on direct remote physiological influ-ences

Note that in such physiological blocking studies the influence attempt isdefined as unwelcome only in terms of the volition of the potential influenceethe influence attempt is not itself intended to be harmful but it is the intentionof the influencee to resist such influence nevertheless This paradigm enablesan indirect approach to the question of harmful psi and it also allows us to ex-plore more theoretical questions as to the limits of remote mental influenceWe conducted two studies that were conceptual replications and extensions ofBraud and Schlitzrsquos psi-blocking study The two studies had very similarmethodologies to one another so a detailed description of the method forStudy 1 will be given and we will highlight the differences between the studieswhen we introduce Study 2

Study One Method

The general procedure followed the standard direct mental influence withliving systems (DMILS) method as outlined by Braud and Schlitz (1991) Theinfluencee (participant) was sensorially isolated from the influencer (experi-menter) The influencer monitored the influenceersquos EDA in real time via acomputer display meanwhile following a random influence schedule given bythe computer The schedule told the influencer when to activate calm or haveno influence (rest) on the influenceersquos physiology The influencee was blind tothe instructions for the influence schedule The influencer was blind to the in-fluenceersquos instructions We also administered a vividness of mental imageryquestionnaire and recorded qualitative information on participantsrsquo mentalstrategies to gain some insight into the psychological aspects of resisting a re-mote influence attempt

Blocking vs Co-operating 517

518 C Watt et al

Design

We used a within-subjects design (whereas Braud and Schlitz used a between-subjects design with different groups of volunteers blocking or co-operating)The independent variable was the influenceersquos strategy (blocking vs co-oper-ating) while the influencer attempted either to calm activate or have no influ-ence over the influenceersquos physiology The dependent variable was the influ-enceersquos physiology EDA being the principal measure but skin temperatureand finger pulse volume being recorded for exploratory purposes (we do notreport these exploratory measures in this paper) Our primary aim was to deter-mine whether there was a smaller remote influence effect when the influenceewas blocking compared to when they were co-operating with the influence at-tempt Our secondary aim was to ascertain whether there was any evidence ofan overall remote influence effect

Influence Schedule

The influence schedule that was followed by the influencer was composedof blocks of four epochs mdash either Rest-Activate-Rest-Calm (RARC) or Rest-Calm-Rest-Activate (RCRA) The ordering of each of these blocks of four wasrandomly determined by the computer as the session proceeded so that in totalthere was an equal number of activate and calm epochs Note that the randomsequence also serves to counter-balance the activatecalm conditions acrossthe duration of the study Each epoch was 20 seconds in duration The experi-mental session was thus split into two runs of equal length each having 33recording epochs ie eight blocks of four (RARC or RCRA) plus an extrarest at the end of each run There was a total of 66 recording epochs over theentire session Between the two runs there was a 150 second rest period duringwhich the influencee and the influencer remained in their rooms

The duration of each epoch and the total number of epochs were chosen inorder to maximize the trade-off between gathering sufficient data and keepingthe duration of the session (2 runs of 10 min) at a comfortable length for par-ticipants Additionally we felt that 20 second epochs would be sufficient toenable any stimulus-response effect to manifest in EDA A liberal strategy(that used by Radin et al 1995) was adopted to specify the criteria for includ-ing data if a session was aborted due to technical faults with the psychophysi-ology apparatus data was included from any session that had at least one calmand one activate period (ie a total of four recording epochs)

Data Treatment

Braudrsquos Percentage Score Index (PSI as described in Braud amp Schlitz1991) per session was the unit of analysis used in this study PSI is a conserva-tive summary of the results of an entire session and is calculated by dividingthe sum of EDA activity over all Activate epochs by the sum of all calm andactivate EDA [PSI = S A( S C+ S A)] The rest epochs were not used in this

analysis Their function was primarily to separate the calm and activateepochs from one another and to separate any lingering influence that may havebeen left at the end of the influencerrsquos calm or activate epochs In the absenceof any remote influence effect PSI equals 50 A PSI of greater than 50 rep-resents a remote influence effect in the predicted direction (ie calming whenthat was the influencerrsquos intention activating when that was the influencerrsquosintention) A PSI of less than 50 represents a remote influence effect in thedirection opposite to that predicted For the present study we calculated PSIseparately for each participantrsquos blocking and co-operating condition plus wecalculated an overall PSI score for each participant

Participants

The influencees were volunteer participants drawn from the authorsrsquo con-tacts in the Department of Psychology at Edinburgh University and includingundergraduate psychology students recruited by e-mail and by word of mouthThere were three experimentersinfluencers overall (Watt Ravenscroft andMcDermott) but in each testing session there were only two individuals mdash thevolunteer influencee and the influencer Each experimenter also took part asinfluencer in one session

Written Materials

Each volunteer was given a one page information sheet Once they had readthis they were given a 34-item participant information form (PIF) and a 16-item Vividness of Visual Imagery Questionnaire (Marks 1973) After dis-cussing the procedure with the influencer each participant was asked to sign aconsent form that assured confidentiality and noted their right to withdrawfrom the study at any time Following completion of the DMILS procedure in-fluencee and influencer both completed one-page qualitative post-sessionquestionnaires in which they were invited to describe the kinds of strategiesthat they used during the testing session and to give their subjective impres-sions of the session A letter of thanks and a brief overview of the results weregiven to participants at the end of the study

Technical Apparatus

Two rooms were used for the testing sessions The influenceersquos room wassound-attenuated with double-doors and was located approximately 20 metersaway from the influencerrsquos room down a short flight of stairs The DMILS ap-paratus (an I-410 multichannel physiological monitoring unit from J amp J Sys-tems with source code from Physiodata) was set up between these two roomsand was run on two time-synchronized Pentium computers In the influenceersquosroom a monitor displayed written instructions and a mildly engaging visualdisplay (a screen saver shown during the actual data collection epochs) to theinfluencee This room also contained speakers that gave auditory instructions

Blocking vs Co-operating 519

520 C Watt et al

(ldquohey block nowrdquo ldquohey co-operate nowrdquo and ldquosession over thank yourdquo) tothe influencee just prior to each of the two 10-minute runs duplicating thewritten instructions that appeared on the computer monitor The physiologysensors that were attached to the influenceersquos fingers were located in the influ-enceersquos room The signals from the sensors were relayed along a cable andprocessed by a computer located in the influencerrsquos room Therefore the com-puter in the influenceersquos room was merely used to provide instructions and avisual display to the influencee The computer in the influencerrsquos room con-trolled the DMILS software and processes and stored the psychophysiologydata

Procedure

The influencer and participant initially sat down in the reception room in theparapsychology suite for a further description of procedure and returned theircompleted PIFimagery questionnaire to the influenceer who went over it withthe participant Any questions that were raised by the participant prior to test-ing were answered here (This question-and-answer session usually lasted fif-teen minutes) The participant then gave their informed consent to the studyThe influencer also suggested various mental approaches the participant mightadopt for the blocking and co-operating conditions (some examples weregiven in the information sheet) These suggestions included both visual im-agery (eg visualizing a shielding cocoon in the blocking condition) and cog-nitive strategies (eg adopting a stubborn and uncooperative frame of mind inthe blocking condition) Participants were informed that different strategiesmight suit different people and they were to do whatever felt natural for themThe influencer then took the participant to the testing room and demonstratedthe auditory instructions so that the participant would not be startled by thesewhen they occurred during the session itself The participant was also taken tosee where the influencer was sitting and the influencer described what theywould be doing when they were attempting to activate or calm the participant

Once the preliminaries were finished the influencee was seated in a partiallyreclined chair in front of the computer monitor and the physiology sensorswere attached to their fingers by the influencer The room was softly illuminat-ed by a desk lamp with an orangered filter At this stage the influencer checkedthat the physiology signals were being transmitted to the computer in theirroom Returning to the influencee the influencer set a start time on the influ-enceersquos computer This was on average two to three minutes later The influ-encer then returned to their room closing all doors and waited for the start ofthe session

Physiology Data Collection As the session began the influencee was in-structed visually and auditorially either to ldquohey block nowrdquo or ldquohey co-oper-ate nowrdquo The influenceersquos monitor then displayed a colorful and dynamicscreen-saver that was intended to be engaging without being particularlyarousing or relaxing At the same time the influencer was following the influ-

ence schedule as instructed by the computer observing the real-time feedbackdisplay of the participantrsquos EDA when desired At the end of the first run whichlasted approximately ten minutes the influencer saw a 150 second count downon the computer before the second run began At the same time the influenceesaw a message about the strategy to adopt for the forthcoming second run (ieto block or to co-operate whichever strategy was not used for the first run) Atthe beginning of the second run the influencee was also instructed auditoriallyeither to ldquohey block nowrdquo or ldquohey co-operate nowrdquo The dynamic visual dis-play resumed while the influencer followed a random influence schedule forapproximately ten minutes The entire session took about 25 minutes to com-plete At the end of the session the influencee was auditorially and visually in-formed ldquosession over thank yourdquo and waited until the influencer returned tothe influenceersquos room

Post-Testing Procedures The influenceersquos physiology sensors were re-moved by the influencer and both returned to the reception room to completethe post-session questionnaires Then the influencee was taken to the influ-encerrsquos room to see feedback on the results of their testing session There wasan option for some final discussion about the session after which the volun-teer was thanked for their participation and left Once all the data was collect-ed feedback on the results of the study as a whole was sent to participants

Planned Hypotheses and Analyses

Our main aim was to examine whether there was any evidence for blockingof a remote mental influence on physiology [EDA is the principal physiologymeasure here with per-session PSI scores representing the measure of theDMILS effect on EDA] H1 therefore predicted significantly greater PSIscores in the co-operating condition than in the blocking condition (using a re-lated t-test one-tailed) Our second question was whether there was any over-all evidence of remote mental influence on EDA This analysis would be con-ducted on the data of blocking and co-operating runs combined if nosignificant difference was found between these two conditions (as tested byH1) If H1 was supported H2 would be tested on the data of co-operating runsonly predicting PSI scores significantly greater than chance expectation(using a single-mean t-test of the PSI score vs MCE one-tailed)

Planned alpha level was 05 There was no correction for multiple analysisThis was dealt with by restricting the number of formal hypotheses (with one-tailed significance tests) to two in each study All other analyses were regardedas exploratory and two-tailed significance tests were used

Exploratory Measures

Variance

DMILS researchers do not yet appear to have identified what might be themost appropriate indicator of a remote influence effect Therefore in addition

Blocking vs Co-operating 521

522 C Watt et al

to the PSI score we looked at the amount of variance that was shown in block-ing and co-operating scores The variance measure was calculated as followsMean values for activate periods and calm periods were calculated separatelyfor block and co-operate conditions Mean values were taken from each indi-vidual score in that condition and squared and summed S [mean value for con-dition - individual score]2 This gave a summation of variability around themean for each activate calm and rest period for block and cooperate condi-tions It was used in conjunction with the mean level of EDA activity givingan indication of the motion of the wave (ie the variability of the EDA activi-ty) rather than the overall level

VVIQ

Anecdotally the use of appropriate imagery (eg visualize being surroundedby a shielding white halo of white light) is suggested to be an effective blockingtechnique (eg Hope 1983) We therefore wished to see if participants who re-ported experiencing vivid imagery with ease are perhaps more effective block-ers than those who do not report clear mental imagery Descriptive data wasgathered to explore any trends in PSI scores for blockingco-operating forthose participants with VVIQ scores ranging from low (ie reporting vividmental imagery) to high (ie reporting little or no mental imagery)

Qualitative Reports

We also wanted to report data illustrating what kind of different strategiesparticipants used when they were blocking and co-operating These qualitativereports were gathered in the form of the participantsrsquo written responses toopen-ended questions on a post-session questionnaire Watt and Ravenscroftindependently coded these written responses into categories and cross-checked to ensure coding agreement

Sex and Experimenter Effects

We also examined the data for sex effects and experimenter effects becausethese variables could confound the interpretation of any overall effects Final-ly a post hoc analysis was conducted on influenceeinfluencer sex pairings Ina study on the unconscious detection of remote staring Schlitz and LaBerge(1997) found the greatest effect for opposite-sex observer-subject pairs

Study One Results and Discussion

Participants

Thirty-two individuals took part as influencees (12 females 20 males) theirages ranged from 19 to 63 years (mean age 285 years)

Incomplete Data

One participant declined to give their age on the PIF form so the descriptivestatistics on age do not include this participant (so N = 31) Also due to a tech-nical hitch the exploratory variance scores were lost for one participant so N = 31 for this measure There were no drop-outs from this study

Remote Influence Measure

Table 1 presents a summary of the remote influence data in the form ofmean Percentage Score Index for co-operating and blocking conditions Therewas no significant difference between PSI for co-operating and blocking con-ditions (t = - 202 df = 31) Therefore Hypothesis 1 (predicting significantlygreater PSI scores for co-operating condition compared to blocking condition)was not supported by these results As planned the PSI scores for co-operatingand blocking runs were combined to examine Hypothesis 2 (of an overall re-mote mental influence effect in line with the calming and activating intentionsof the influencee) Mean PSI scores were exactly at chance expectation (t = - 031 df = 31) therefore there was no support for Hypothesis 2

Exploratory Questions

Variance Measure

Table 2 presents summary data for the exploratory measure of the amount ofvariance shown in each condition (lt 50 represents reduced variance gt 50 rep-resents increased variance) There was no significant difference between vari-ance scores for co-operating and blocking conditions (t = 835 df = 30) andwhen the conditions were combined variance did not depart from chance ex-pectation (t = - 244 df = 30) It is interesting to note however that the vari-ance measure gives a much wider range of scores than the PSI measure

Blocking vs Co-operating 523

TABLE 1Summary Data Mean PSI Scores

(N = 32 MCE = 50 range in brackets)

Co-operate PSI Block PSI Combined PSI

4996 5003 5000(4692- 5215) (4653- 5404) (4853- 5309)

TABLE 2Summary Data Mean Variance Scores(N = 31 MCE = 50 range in brackets)

Co-operate Variance Block Variance Combined Variance

5089 4864 4977(2648- 7012) (2492- 6484) (4116- 6297)

524 C Watt et al

suggesting that the variance measure may potentially be a more sensitive indi-cator of remote influence effect than PSI

Vividness of Mental Imagery

Low VVIQ scores represent participants reporting highly vivid imagery andhigh VVIQ scores represent those reporting little or no imagery The meanscore was 4275 (range from 21 to 80) Using a Pearson correlation we foundthat there was no evidence to suggest that those participants with low VVIQscores (high imagery) were better blockers than those with high scores (VVIQvs PSI blocking r = - 113) We also found that there was no evidence of a cor-relation between vividness of imagery and overall remote influence effectsusing the combined Percentage Score Index measure (VVIQ vs PSI overall r =- 077)

Influenceesrsquo Blocking and Co-operating Strategies

Influencees completed a questionnaire (after the session was over) askingthem to describe what strategies they used for blocking and co-operating Cat-egorizing these responses we found a difference in the kinds of strategies usedwhen blocking compared to co-operating as shown in Table 3 In the blockingcondition the majority of participants (469) used visual imagery strategiesFewer (281) participants used a combination of imagery and cognitivestrategies (such as repeating words related to blocking) while 25 reportedusing purely cognitive strategies for blocking (such as adopting a resistant anduncooperative attitude or thinking of irrelevancies) In the co-operating con-dition fewer participants used visual imagery alone (333) The majority ofparticipants (424) reported that their strategies for co-operating includedadopting a feeling of passiveness openness or relaxation whereas no partici-pants adopted this kind of strategy for the blocking condition 121 of partic-ipants used cognitive strategies alone (such as adopting a co-operative atti-tude) and the same proportion used a combination of visual imagery andcognitive strategies

Therefore there was a tendency for participants to use visual imagery forblocking and relaxationpassivity or visual imagery for co-operating This wasdespite the fact that we asked participants to attempt to adopt similar levels ofactivation for both co-operating and blocking strategies There was also a ten-dency not surprisingly for those participants with low VVIQ scores (ie vivid

TABLE 3Influenceesrsquo Mental Strategies ()

Vis Im Cognitive VisCog OpenRelax

Blocking 469 250 281 00Co-operating 333 121 121 424

imagers) to use visual imagery strategies Those who had high VVIQ scoresalso used visual imagery but more often used this in combination with othertactics

Other Qualitative Reports from Participants According to their post-ses-sion questionnaire responses participants generally felt comfortable about theattempt to be influenced (N = 24) Three participants said they felt apprehen-sive or uncomfortable with the idea of being influenced Interestingly of thesethree two were actively involved with parapsychology and were aware of theharmful psi implications discussed at the beginning of this report The thirdwas a psychology student who had attended a lecture given by Watt in whichthe possibilities of harmful psi were discussed Another group of three partici-pants said they would be uncomfortable with the idea of being influenced ifthey believed such influence was possible

Seven participants said they found it was difficult to sustain concentrationfor the 10 minutes that comprised each condition Five participants found thescreensaver distracting or disturbing Two participants specifically mentionedthat they found blocking a more taxing task than co-operating

Sex Effects

There were more males (N = 20) than females (N = 12) in this study howev-er there was no significant difference in PSI scores for remote physiology in-fluence between males and females (male PSI vs female PSI t = 1085 male PSI = 4987 female PSI = 5021)

Experimenter Effects

We studied whether there was any difference in PSI scores obtained by thethree experimenters The results are shown in Table 4 There seems to be nosign of experimenter effects in study one

InfluenceeInfluencer Sex Pairings

Table 5 gives the mean PSI scores for different influenceeinfluencer pair-ings Taking the combined PSI scores it can be seen that the highest remote in-fluence effect is for FM pairs followed by MF pairs followed by FF pairsand the lowest scoring is for MM pairs Post hoc it was found that the mixedsex pairings showed significantly higher remote influence scores than the same

Blocking vs Co-operating 525

TABLE 4Remote Influence Results by Experimenter

Experimenter N sessions Coop PSI Block PSI Combined PSI

CW 12 4983 5011 4997ZMcD 10 5015 4998 5006JR 10 4994 4999 4996

526 C Watt et al

sex pairings (t = - 2016 df = 30 p = 053 2 - t) This replicates the effect foundby Schlitz and LaBerge (1997)

Study One Conclusions

In Study 1 we developed a methodology to examine intentional resistance toa remote mental influence attempt (Watt et al 1997) We contrasted influ-enceesrsquo physiological activity when they were mentally co-operating with aremote attempt to influence their physiology with physiological activity wheninfluencees were mentally blocking the influence attempt We found no evi-dence for any remote influence effect overall and no difference betweenblocking and co-operating conditions Our qualitative measures yielded someinteresting information about participantsrsquo mental strategies and reactions dur-ing the remote influence session We found no difference between experi-menters in remote influence scores but we did find that mixed sex influ-enceeinfluencer pairings scored more highly than same sex pairings

Study Two Method

Study 2 was planned to explore in more depth the nature of the relationshipbetween aspects of personality and cognition and psi performance Partici-pants took part in three testing sessions approximately one week apart Thefirst consisted of various psychological measures the second consisted of aforced choice ESP task and the third consisted of a replication of the blocking-co-operating study (Study 1) described above The psychological measureswere not scored until after completion of the remote influence session Partici-pants received feedback about their forced choice ESP performance at the con-clusion of the second session For ease of interpretation the present paper isrestricted to a description of the third testing session which was essentially areplication of study one with fewer influencers and greater statistical power

We introduced some slight changes to the methodology of Study 2 based onour experiences with Study 1 At the start of each of the two runs in Study 1the participant was given auditory instructions either to ldquoblock nowrdquo or ldquoco-operate nowrdquo We observed that the auditory instructions appeared to affect theparticipantrsquos EDA as if they were startled As planned the first recordingepoch in each run was not used in analysis to give the participantrsquos EDA time

TABLE 5Mean PSI Scores for Different InfluenceeInfluencer Sex Pairings

Pairing N Coop PSI Block PSI Combined PSI

M-F 6 4932 5069 5001F-M 6 5043 5034 5048F-F 6 5022 4952 4993M-M 14 4988 4975 4981

to settle down However in order to remove any remaining potential artifactu-al effects on EDA in Study 2 only visual instructions were given

Another difference in methodology concerned the screensaver viewed bythe participants during the experimental session In Study 1 some participantsreported that the series of patterns produced by the computer were being dis-played at too fast a rate such that some participants found it distracting or dis-turbing Again we felt that this may have an artifactual effect on EDA so wedecided for study two to reduce the speed of the screensaver

Thirdly in Study 2 we did not give participants detailed information abouttheir remote influence performance until the feedback letters that were distrib-uted at the conclusion of the entire study The reason for this change was mere-ly to streamline the procedure on what was already a time-consuming experi-ment Finally in another streamlining change we did not take VVIQ measuresin Study 2

Planned Hypotheses and Analyses

These were the same as with Study 1 That is hypothesis 1 predicted higherPSI scores for the co-operating condition than for the blocking condition(measured with a related t-test one-tailed) If no significant difference wasfound between blocking PSI and co-operating PSI scores the two would becombined to look at overall remote influence effect (hypothesis 2 single-meant-test one-tailed) If a significant difference was found with hypothesis 1 theanalysis for hypothesis 2 would be conducted with the data for co-operatingsessions only

Exploratory Questions

As with study one we took an exploratory variance measure of remote EDAinfluence We also gathered qualitative descriptions of the mental strategiesparticipants adopted whilst blocking and co-operating We also looked at sexeffects experimenter effects and post hoc influenceeinfluencer sex pairings

Participants

Study 2 had two experimenterinfluencers (Watt amp Ravenscroft) After aparticipant was recruited (usually by Ravenscroft) they were assigned to oneof the two experimenters for all three testing sessions This assignment wasdone unsystematically according to the preferences and availability of the ex-perimenters As planned 50 volunteer participants were drawn from the gener-al participant pool at the Koestler Parapsychology Unit and by word ofmouth All participants completed the Participant Information Form and aConsent form prior to the study The participants consisted of 33 females and17 males Average age was 41 years range from 22 years to 76 years of age Inorder to increase the studyrsquos relevance to the general population undergradu-ate psychology students were not recruited as participants The main criterion

Blocking vs Co-operating 527

528 C Watt et al

for participation was that the participants were interested in and open-mindedabout parapsychology Again there were no drop-outs from this study Allparticipants received a detailed personal feedback letter at the conclusion ofthe study

Study Two Results and Discussion

Remote Influence Measure

PSI scores are summarized in Table 6 which shows that the mean PSI scorewhen co-operating was nonsignificantly lower than mean PSI score whenblocking (t = - 595 df = 49) Hypothesis 1 was therefore not supported Asplanned in the absence of any significant difference between blocking and co-operating PSI scores the two conditions were combined to test hypothesis 2which predicted overall significant remote mental influence performance Thecombined PSI score was significantly greater than mean chance expectation of50 (t = 1806 df = 49 p = 038 1 - t) Therefore hypothesis 2 was supportedgiving evidence of an overall remote influence effect upon psychophysiologyin the prespecified direction

The failure to support hypothesis 1 confirms what we found in Study 1 thatthere was no significant difference between blocking and co-operating condi-tions on remote influence scores Indeed both studies found a slight effect inthe opposite direction to that predicted that is slightly higher PSI scores whenblocking was the influenceersquos intention than when co-operating was their in-tention It was encouraging to find significant support for a remote influenceeffect in the predicted direction for blocking and co-operating conditions com-bined No such trend was found in the previous study though with a smallernumber of participants (N = 32) the earlier study had lower statistical powerthan the present study

Exploratory Questions

Variance Measure

Table 7 shows the results for the exploratory measure of EDA variance lt 50represents reduced variance gt 50 represents increased variance We found nodifference between variance scores for cooperate versus block conditions(t = 150 df = 49) When these are combined the resulting figure does not de-viate from chance expectation (t = 440 df = 49) Thus as in Study 1 it seems

TABLE 6Summary Data Mean PSI Scores

(N = 50 MCE = 50 range in brackets)

Co-operate PSI Block PSI Combined PSI

5008 5019 5016(4821- 5221) (4807- 5557) (4900- 5296)

that the exploratory variance measure does not reveal any additional indica-tion of the operation of a remote mental influence in this study However thewider range of variance scoring continues to suggest that this measure mayhave greater potential sensitivity than the summary PSI score

Influenceesrsquo Blocking and Co-operating Strategies

We were interested in learning more about what particular strategies partici-pants chose to adopt when blocking and co-operating As in Study 1 a post-session questionnaire was used to gather this information Table 8 shows thepercentage of participants who adopted each kind of strategy

The table shows a very similar pattern to that found in Study 1 The mainpoint of interest is that almost no participants took a passive approach toblocking (ie adopting an openrelaxing strategy) A much higher proportionpreferred to adopt this approach when co-operating The preferred mode forboth conditions was to use some form of visual imagery such as imagining ashield or a screen when blocking and imagining an open channel such as ariver when co-operating

The majority of participants reported no concerns over the possibility ofhaving their physiology remotely influenced by another person as found instudy one

Sex Effects

As in study one the sexes were not equally represented among participantsWhereas Study 1 had an excess of male participants in Study 2 twice as manyfemales (N = 33) as males (N = 17) participated It was therefore important tolook for sex differences since any such differences could confound the inter-pretation of overall trends We found no difference between sexes when co-op-erating (female mean PSI score = 5008 male mean PSI score = 5006 t = 107 df = 48) There was a nonsignificant trend for females to have higher

Blocking vs Co-operating 529

TABLE 7Summary Data Mean Variance Scores(N = 50 MCE = 50 range in brackets)

Co-operate Variance Block Variance Combined Variance

5076 5054 5034(3164- 7279) (3572- 6813) (3749- 6671)

TABLE 8Influenceesrsquo Mental Strategies ()

Visual Imagery Cognitive VisCog OpenRelax

Blocking 64 26 6 2Co-operating 60 8 12 20

530 C Watt et al

PSI scores than males when blocking (mean PSI scores 5030 and 4998 re-spectively t = 992 df = 48) Combining blocking and co-operating conditionsthere was no overall difference between the sexes (mean PSI scores 5020 forfemales 5008 for males t = 665 df = 48)

Experimenter Effects

Study 2 found a statistically significant overall remote influence effect(combined PSI = 50155 t = 1806) Table 9 shows the breakdown of remoteinfluence scores for the two experimenters CW and JR Looking at the overalldata for the two experimenters we see that CWrsquos sessions were associatedwith most of the remote influence effect However due to the smaller N CWrsquossessions are not independently significant (t = 1632) and there is not a signifi-cant difference between CW and JRrsquos overall PSI scores (t = 678) Post hocwe calculated effect size r according to the formula r = [t2(t2+df)]12 for the twoexperimenters For JR (t = 817 df = 22) we found a small effect sizer = +17 For CW (t = 1632 df = 26) the effect size was moderate r = +30This may be some post hoc indication of an experimenter effect in Study 2

InfluenceeInfluencer Sex Pairings

As can be seen from Table 10 Study 2 found exactly the opposite trend tothat found in Study 1 On a post hoc test same sex pairings scored significant-ly higher on the remote influence PSI measure than mixed sex pairings(t = - 2034 df = 48 p = 048 2 - t) This contradicts Schlitz and LaBergersquos(1997) finding of greater autonomic detection of remote staring for mixed sexpairs

TABLE 9Remote Influence Results by Experimenter

Experimenter N sessions Coop PSI Block PSI Combined PSI

CW 27 5011 5026 5021JR 23 5003 5010 5009

TABLE 10Mean PSI Scores for Different InfluenceeInfluencer Sex Pairings

Pairing N Coop PSI Block PSI Combined PSI

MF 11 4977 5009 4997FM 17 4983 5021 5003FF 16 5035 5038 5038MM 6 5060 4978 5027

General Summary and Conclusions

We have described the findings of two studies exploring the limits of remotemental influence effects We wished to see whether the mental strategiesadopted by the influencee would affect remote influence scores In particularcould an influencee block or resist a remote influence attempt We recorded in-fluenceesrsquo electrodermal activity while the influencer attempted to calm to ac-tivate or to have no influence over the influenceesrsquo physiology In each studythere were two conditions reflecting the influenceesrsquo designated mental strate-gies in response to the remote influence attempt blocking and co-operating

There were only two formal hypotheses 1) that remote influence scoreswould be significantly higher during the co-operating condition than duringthe blocking condition and 2) that overall there would be significant evidenceof remote mental influence on physiology in the designated direction (iecalming when calming was the influencerrsquos intention activating when activat-ing was the influencerrsquos intention) Contrary to prediction Study 1 found thatremote influence effects were higher in the block condition than in the co-op-erate condition (but not to a significant extent) Also overall remote influenceperformance was at chance level This study therefore found no support for ei-ther of the formal hypotheses Study 2 consisted of a replication of Study 1with a few minor adjustments plus additional measures that were not reportedin this paper Again contrary to prediction remote influence scores were non-significantly higher in the block condition than in the co-operate conditionThis replicates the trend found in Study 1 failing to support hypothesis oneThese results suggest that the mental strategy adopted by the influencee in theface of a remote influence attempt has no effect on remote influence perfor-mance Referring back to the questions posed in the introduction then wehave not found any evidence to suggest that it is possible to resist a remote in-fluence effect and there is as yet no indication of the possible limiting condi-tions for remote mental influence

Unlike Study 1 Study 2 found statistically significant evidence of an overallremote influence effect in the prespecified direction (ie calming of EDAwhen that was the influencerrsquos intention activation of EDA when that was theinfluencerrsquos intention) In order to facilitate comparison with a recent meta-an-alytic review of remote influence studies we calculated the effect size (r) ac-cording to the formula r = [t2(t2+df)]12 For their review of 19 EDA influencestudies Schlitz and Braud (1997) found an average effect size of r = +25 OurStudy 1 with overall t = - 031 df = 31 gives a negligible effect size of r = - 006 Study 2 with overall t = 1806 df = 49 gives r = +25 Study 2 there-fore replicates the mean effect size found in EDA studies conducted by otherresearchers in other parapsychology laboratories Post hoc we can combinethe data of the two studies for greater statistical power Overall t = 1176 df = 81 giving a small effect size of r = +13

We can find no clear reason why Study 2 found an EDA influence effectwhile Study 1 did not A number of factors are worth considering 1) The

Blocking vs Co-operating 531

532 C Watt et al

significant result could be due to chance or undetected artifact 2) Study 2 im-plemented what we considered to be minor methodological improvements thatmight have removed some artifactual fluctuations from the EDA data that weobserved in study one 3) Each experimenter in Study 2 conducted three testingsessions with their participant By the third session in which the remote influ-ence measure was taken experimenters were on quite friendly terms with theirparticipants perhaps more so than in Study 1 where each participant came infor only one testing session 4) The participants in Study 2 were more represen-tative of the general population we felt than those in Study 1 Study 1 in-volved more undergraduate psychology students and associates of the experi-menters 5) The experimenters in study one were less experienced with theremote influence protocol than when they conducted Study 2

In addition to the formal hypotheses we reported a number of exploratorymeasures Study 1 found that influenceesrsquo self-reported vividness of visual im-agery appeared to be unrelated to the remote influence performance In bothstudies influenceesrsquo qualitative descriptions of the actual mental strategiesthey employed when blocking and co-operating revealed a tendency for differ-ent strategies to be preferred for the different conditions The use of visual im-agery was popular for both blocking (eg visualizing being surrounded by aprotective cocoon) and co-operating (eg visualizing a river flowing betweeninfluencer and influencee) However a considerable proportion of participantsreported adopting an open or relaxed frame of mind when co-operating whilstalmost no participants chose this strategy when blocking This was despite thefact that prior to each session we encouraged participants to adopt comparablestrategies for blocking and co-operating

It was interesting and perhaps surprising to note that in both studies themajority of participants reported no feelings of concern or discomfort at thepossibility of having their physiology remotely influenced Those who wereconcerned tended to be those who knew more about parapsychology and thepossible harmful applications of remote influence A small proportion of par-ticipants was unconcerned because they did not believe such an influence waspossible in the first place

Both studies employed an exploratory variance measure of EDA influenceThis measure showed a greater range of scoring overall thus perhaps greaterpotential sensitivity to a remote influence effect However the variance mea-sure showed no indication of a remote influence effect To date researchershave employed a variety of indicators of remote influence as the literaturegrows progress may be made on identifying which is the most appropriate in-dicator

Neither study found significant sex differences in remote influence effectsthough in both studies female influencees tended to have higher remote influ-ence scores than male influencees

In terms of their overall remote influence effects there appeared to be nodifference between experimenterinfluencers in Study 1 (r = +06 - 05 and

- 06) In Study 2 effect size measures indicated that the experimenters dif-fered in remote influence effect sizes (r = +17 and +30) Due to the smaller Nthis difference was not found to be statistically significant Post hoc combin-ing Studies 1 and 2 for experimenters JW and CW (ZMcD was not an experi-menter in Study 2) the effect sizes are r = +08 and r = +20 respectively

We conducted a post hoc analysis of the effects of influenceeinfluencer sexpairings on remote influence scores Our two studies came up with contradic-tory findings each statistically significant Study one found that mixed sexpairings had significantly higher remote influence scores than same sex pair-ings This replicated the post hoc finding of Schlitz and LaBerge (1997) in con-nection with the autonomic detection of remote observation Study 2 foundthat same sex pairings had significantly higher remote influence scores thanmixed sex pairings Our contradictory findings do not help to clarify the pic-ture on the importance or otherwise of sex pairings It is quite possible that sexis not the most salient aspect of the dynamic between pairs Age personalityand familiarity are likely to be important variables moderating the relationshipbetween influencer and influencee

There are many more variables that we could look at but as the number ofanalyses multiplies so does the opportunity for spurious significant resultsFor this reason we have named only two formal hypotheses for each experi-ment All other analyses are regarded as exploratory and any trends that wehave highlighted above would require replication

In conclusion these two studies have found no evidence of a difference inremote influence effects when influencees are asked to adopt cooperative men-tal strategies compared to when they are asked to adopt blocking mental strate-gies In fact contrary to expectation both studies found slightly higher remoteinfluence effects for the blocking condition compared to the co-operating con-dition On this basis we have found no indication of limiting conditions for re-mote influence effects Therefore while anecdotal advice exists on how to de-fend oneself against a remote physiological influence attempt in the realworld laboratory evidence to date has provided no support for this advice

Given the importance of this question both in practical terms and in theoret-ical terms we argue that this line of research deserves continued investigationIf a way could be found to conduct research with a more realistic scenario mdash where the remote influence was not just neutrally and abstractlydefined as unwelcome mdash then our research might have more meaningful thingsto say about the parameters of remote influence It is interesting and perhapssurprising to note that when Watt has lectured on this topic to informed lay au-diences the feedback has often been that researchers must use more potential-ly harmful or disruptive remote influence (of course with the informed consentof the participants) While there are clear ethical inhibitions when consideringpossible physiological influence perhaps research using higher level influencewould be more acceptable For example one could attempt to disrupt or en-hance the concentration of chess players who were aware of the influence

Blocking vs Co-operating 533

534 C Watt et al

attempt and who were concurrently blocking or co-operating with the influ-ence attempt Research already exists to suggest that remote influence does notjust operate at physiological levels Attention focusing for example can be fa-cilitated by a remote helper (Brady amp Morris 1997 Braud et al 1995) Futureresearch in this area must therefore expand beyond physiological influences

Acknowledgements

The authors gratefully acknowledge the financial support they receivedfrom the Institut fuumlr Grenzgebiete der Psychologie und PsychohygieneFreiburg Germany while conducting these studies We would also like tothank Dr Paul Stevens for generously helping to solve software difficultieswhen preparing for these studies and the anonymous reviewer for helpfulcomments Address reprint requests to Dr C Watt Department of PsychologyUniversity of Edinburgh 7 George Square Edinburgh EH8 9JZ Scotland UK

ReferencesBrady C amp Morris R L (1997) Attention focusing facilitated through remote mental interac-

tion A replication and exploration of parameters In R Wiseman (Ed) Proceedings of Pre-sented Papers The Parapsychological Association 40th Annual Convention held in conjunc-tion with The Society for Psychical Research pp 73ndash91

Braud W G (1985) Blockingshielding psychic functioning through psychological and psychictechniques A report of three preliminary studies In RA White and J Solfvin (Eds) Re-search in Parapsychology 1984 Metuchen NJ Scarecrow Press pp 42ndash44

Braud W G amp Schlitz M J (1991) Consciousness interactions with remote biological systemsAnomalous intentionality effects Subtle Energies 2 1

Braud W G Schlitz M J Collins J and Klitch H (1985) Further studies of the bio-PK effectFeedback blocking specificitygenerality In RA White amp J Solfvin (Eds) Research inParapsychology 1984 Metuchen NJ Scarecrow Press pp 45ndash48

Braud W G Shafer D McNeill K and Guerra V (1995) Attention focusing facilitatedthrough remote mental interaction Journal of the American Society for Psychical Research89 103

Cohen J (1997) Statistical Power Analysis for the Behavioral Sciences New York AcademicPress

Dalton K (1994) Remotely influenced ESP performance in a computer task A preliminarystudy Proceedings of the 37th Annual Convention of the Parapsychological Association pp95ndash103

Dossey L (1994) Healing and the mind Is there a dark side Journal of Scientific Exploration 873

Halifax-Grof J (1974) Hex death In A Angoff amp D Barth (Eds) Parapsychology and Anthro-pology Proceedings of an International Conference Held in London England August 29ndash311973 NY Parapsychology Foundation

Hope M (1983) Practical Techniques of Psychic Self-Defence Wellingborough Northampton-shire Aquarian Press

Marks DF (1973) Visual imagery differences in the recall of pictures British Journal of Psy-chology 64 17

Nash CB (1982) Psychokinetic control of bacterial growth Journal of the Society for PsychicalResearch 51 217

Radin DI Taylor RK and Braud WG (1995) Remote mental influence of human electroder-mal activity A pilot replication European Journal of Parapsychology 11 19

Schlitz M amp Braud W G (1997) Distant intentionality and healing Assessing the evidence Al-ternative Therapies 3 62

Schlitz M J amp LaBerge S (1997) Covert observation increases skin conductance in subjectsunaware of when they are being observed A replication Journal of Parapsychology 61 185

Schmeidler G R (1958) Agent-percipient relationships Journal of the American Society forPsychical Research 52 47

Schmeidler G R (1961) Evidence for two kinds of telepathy International Journal of Parapsy-chology 3 5

Tart C T (1979) A survey of expert opinion on potentially negative uses of psi United Statesgovernment interest in psi and the level of research funding of the field In WG Roll (Ed) Re-search in Parapsychology 1978 Metuchen NJ Scarecrow Press pp 54

Tart C T amp Labore C M (1986) Attitudes towards strongly functioning psi A preliminary sur-vey Journal of the America Society for Psychical Research 80 163

Tart C T (1986) Psychicsrsquo fears of psychic powers Journal of the American Society for Psychi-cal Research 80 278

Watt C A McDermott Z amp Ravenscroft J (1997) Exploring the limits of remote mental inter-action A comparison of blocking versus co-operating strategies In R Wiseman (Ed) Pro-ceedings of the 40th Annual Convention of the Parapsychological Association held in conjunc-tion with the Society for Psychical Research Brighton England pp 464ndash476

Blocking vs Co-operating 535

tempted to prevent a remote mental influence by a distant observer upon theirown electrodermal activity A second independent group of sixteen cooperat-ed with the remote influence attempt The influencers were to attempt to in-crease the EDA of the target person They were situated in separate rooms re-mote from the target person but received real-time feedback about the targetpersonrsquos EDA through a visual display The influencers were of course un-aware whether the target person was blocking or co-operating In the blockingcondition participants used imagery of various psi-proof screens shields andbarriers and intended that no remote mental influence would occur Overallthere was no evidence of successful blocking of remote mental influencethough significant blocking was found post hoc for one of the two influencers(Braud 1985 Braud et al 1985) A number of studies have looked at concep-tually related questions of psi-mediated helping (Dalton 1994) and attemptedfacilitation or inhibition of performance at forced-choice ESP tasks (Schmei-dler 1958 1961) and clairvoyance tasks (Braud 1985) However there seemsto have been no follow-up to research on direct remote physiological influ-ences

Note that in such physiological blocking studies the influence attempt isdefined as unwelcome only in terms of the volition of the potential influenceethe influence attempt is not itself intended to be harmful but it is the intentionof the influencee to resist such influence nevertheless This paradigm enablesan indirect approach to the question of harmful psi and it also allows us to ex-plore more theoretical questions as to the limits of remote mental influenceWe conducted two studies that were conceptual replications and extensions ofBraud and Schlitzrsquos psi-blocking study The two studies had very similarmethodologies to one another so a detailed description of the method forStudy 1 will be given and we will highlight the differences between the studieswhen we introduce Study 2

Study One Method

The general procedure followed the standard direct mental influence withliving systems (DMILS) method as outlined by Braud and Schlitz (1991) Theinfluencee (participant) was sensorially isolated from the influencer (experi-menter) The influencer monitored the influenceersquos EDA in real time via acomputer display meanwhile following a random influence schedule given bythe computer The schedule told the influencer when to activate calm or haveno influence (rest) on the influenceersquos physiology The influencee was blind tothe instructions for the influence schedule The influencer was blind to the in-fluenceersquos instructions We also administered a vividness of mental imageryquestionnaire and recorded qualitative information on participantsrsquo mentalstrategies to gain some insight into the psychological aspects of resisting a re-mote influence attempt

Blocking vs Co-operating 517

518 C Watt et al

Design

We used a within-subjects design (whereas Braud and Schlitz used a between-subjects design with different groups of volunteers blocking or co-operating)The independent variable was the influenceersquos strategy (blocking vs co-oper-ating) while the influencer attempted either to calm activate or have no influ-ence over the influenceersquos physiology The dependent variable was the influ-enceersquos physiology EDA being the principal measure but skin temperatureand finger pulse volume being recorded for exploratory purposes (we do notreport these exploratory measures in this paper) Our primary aim was to deter-mine whether there was a smaller remote influence effect when the influenceewas blocking compared to when they were co-operating with the influence at-tempt Our secondary aim was to ascertain whether there was any evidence ofan overall remote influence effect

Influence Schedule

The influence schedule that was followed by the influencer was composedof blocks of four epochs mdash either Rest-Activate-Rest-Calm (RARC) or Rest-Calm-Rest-Activate (RCRA) The ordering of each of these blocks of four wasrandomly determined by the computer as the session proceeded so that in totalthere was an equal number of activate and calm epochs Note that the randomsequence also serves to counter-balance the activatecalm conditions acrossthe duration of the study Each epoch was 20 seconds in duration The experi-mental session was thus split into two runs of equal length each having 33recording epochs ie eight blocks of four (RARC or RCRA) plus an extrarest at the end of each run There was a total of 66 recording epochs over theentire session Between the two runs there was a 150 second rest period duringwhich the influencee and the influencer remained in their rooms

The duration of each epoch and the total number of epochs were chosen inorder to maximize the trade-off between gathering sufficient data and keepingthe duration of the session (2 runs of 10 min) at a comfortable length for par-ticipants Additionally we felt that 20 second epochs would be sufficient toenable any stimulus-response effect to manifest in EDA A liberal strategy(that used by Radin et al 1995) was adopted to specify the criteria for includ-ing data if a session was aborted due to technical faults with the psychophysi-ology apparatus data was included from any session that had at least one calmand one activate period (ie a total of four recording epochs)

Data Treatment

Braudrsquos Percentage Score Index (PSI as described in Braud amp Schlitz1991) per session was the unit of analysis used in this study PSI is a conserva-tive summary of the results of an entire session and is calculated by dividingthe sum of EDA activity over all Activate epochs by the sum of all calm andactivate EDA [PSI = S A( S C+ S A)] The rest epochs were not used in this

analysis Their function was primarily to separate the calm and activateepochs from one another and to separate any lingering influence that may havebeen left at the end of the influencerrsquos calm or activate epochs In the absenceof any remote influence effect PSI equals 50 A PSI of greater than 50 rep-resents a remote influence effect in the predicted direction (ie calming whenthat was the influencerrsquos intention activating when that was the influencerrsquosintention) A PSI of less than 50 represents a remote influence effect in thedirection opposite to that predicted For the present study we calculated PSIseparately for each participantrsquos blocking and co-operating condition plus wecalculated an overall PSI score for each participant

Participants

The influencees were volunteer participants drawn from the authorsrsquo con-tacts in the Department of Psychology at Edinburgh University and includingundergraduate psychology students recruited by e-mail and by word of mouthThere were three experimentersinfluencers overall (Watt Ravenscroft andMcDermott) but in each testing session there were only two individuals mdash thevolunteer influencee and the influencer Each experimenter also took part asinfluencer in one session

Written Materials

Each volunteer was given a one page information sheet Once they had readthis they were given a 34-item participant information form (PIF) and a 16-item Vividness of Visual Imagery Questionnaire (Marks 1973) After dis-cussing the procedure with the influencer each participant was asked to sign aconsent form that assured confidentiality and noted their right to withdrawfrom the study at any time Following completion of the DMILS procedure in-fluencee and influencer both completed one-page qualitative post-sessionquestionnaires in which they were invited to describe the kinds of strategiesthat they used during the testing session and to give their subjective impres-sions of the session A letter of thanks and a brief overview of the results weregiven to participants at the end of the study

Technical Apparatus

Two rooms were used for the testing sessions The influenceersquos room wassound-attenuated with double-doors and was located approximately 20 metersaway from the influencerrsquos room down a short flight of stairs The DMILS ap-paratus (an I-410 multichannel physiological monitoring unit from J amp J Sys-tems with source code from Physiodata) was set up between these two roomsand was run on two time-synchronized Pentium computers In the influenceersquosroom a monitor displayed written instructions and a mildly engaging visualdisplay (a screen saver shown during the actual data collection epochs) to theinfluencee This room also contained speakers that gave auditory instructions

Blocking vs Co-operating 519

520 C Watt et al

(ldquohey block nowrdquo ldquohey co-operate nowrdquo and ldquosession over thank yourdquo) tothe influencee just prior to each of the two 10-minute runs duplicating thewritten instructions that appeared on the computer monitor The physiologysensors that were attached to the influenceersquos fingers were located in the influ-enceersquos room The signals from the sensors were relayed along a cable andprocessed by a computer located in the influencerrsquos room Therefore the com-puter in the influenceersquos room was merely used to provide instructions and avisual display to the influencee The computer in the influencerrsquos room con-trolled the DMILS software and processes and stored the psychophysiologydata

Procedure

The influencer and participant initially sat down in the reception room in theparapsychology suite for a further description of procedure and returned theircompleted PIFimagery questionnaire to the influenceer who went over it withthe participant Any questions that were raised by the participant prior to test-ing were answered here (This question-and-answer session usually lasted fif-teen minutes) The participant then gave their informed consent to the studyThe influencer also suggested various mental approaches the participant mightadopt for the blocking and co-operating conditions (some examples weregiven in the information sheet) These suggestions included both visual im-agery (eg visualizing a shielding cocoon in the blocking condition) and cog-nitive strategies (eg adopting a stubborn and uncooperative frame of mind inthe blocking condition) Participants were informed that different strategiesmight suit different people and they were to do whatever felt natural for themThe influencer then took the participant to the testing room and demonstratedthe auditory instructions so that the participant would not be startled by thesewhen they occurred during the session itself The participant was also taken tosee where the influencer was sitting and the influencer described what theywould be doing when they were attempting to activate or calm the participant

Once the preliminaries were finished the influencee was seated in a partiallyreclined chair in front of the computer monitor and the physiology sensorswere attached to their fingers by the influencer The room was softly illuminat-ed by a desk lamp with an orangered filter At this stage the influencer checkedthat the physiology signals were being transmitted to the computer in theirroom Returning to the influencee the influencer set a start time on the influ-enceersquos computer This was on average two to three minutes later The influ-encer then returned to their room closing all doors and waited for the start ofthe session

Physiology Data Collection As the session began the influencee was in-structed visually and auditorially either to ldquohey block nowrdquo or ldquohey co-oper-ate nowrdquo The influenceersquos monitor then displayed a colorful and dynamicscreen-saver that was intended to be engaging without being particularlyarousing or relaxing At the same time the influencer was following the influ-

ence schedule as instructed by the computer observing the real-time feedbackdisplay of the participantrsquos EDA when desired At the end of the first run whichlasted approximately ten minutes the influencer saw a 150 second count downon the computer before the second run began At the same time the influenceesaw a message about the strategy to adopt for the forthcoming second run (ieto block or to co-operate whichever strategy was not used for the first run) Atthe beginning of the second run the influencee was also instructed auditoriallyeither to ldquohey block nowrdquo or ldquohey co-operate nowrdquo The dynamic visual dis-play resumed while the influencer followed a random influence schedule forapproximately ten minutes The entire session took about 25 minutes to com-plete At the end of the session the influencee was auditorially and visually in-formed ldquosession over thank yourdquo and waited until the influencer returned tothe influenceersquos room

Post-Testing Procedures The influenceersquos physiology sensors were re-moved by the influencer and both returned to the reception room to completethe post-session questionnaires Then the influencee was taken to the influ-encerrsquos room to see feedback on the results of their testing session There wasan option for some final discussion about the session after which the volun-teer was thanked for their participation and left Once all the data was collect-ed feedback on the results of the study as a whole was sent to participants

Planned Hypotheses and Analyses

Our main aim was to examine whether there was any evidence for blockingof a remote mental influence on physiology [EDA is the principal physiologymeasure here with per-session PSI scores representing the measure of theDMILS effect on EDA] H1 therefore predicted significantly greater PSIscores in the co-operating condition than in the blocking condition (using a re-lated t-test one-tailed) Our second question was whether there was any over-all evidence of remote mental influence on EDA This analysis would be con-ducted on the data of blocking and co-operating runs combined if nosignificant difference was found between these two conditions (as tested byH1) If H1 was supported H2 would be tested on the data of co-operating runsonly predicting PSI scores significantly greater than chance expectation(using a single-mean t-test of the PSI score vs MCE one-tailed)

Planned alpha level was 05 There was no correction for multiple analysisThis was dealt with by restricting the number of formal hypotheses (with one-tailed significance tests) to two in each study All other analyses were regardedas exploratory and two-tailed significance tests were used

Exploratory Measures

Variance

DMILS researchers do not yet appear to have identified what might be themost appropriate indicator of a remote influence effect Therefore in addition

Blocking vs Co-operating 521

522 C Watt et al

to the PSI score we looked at the amount of variance that was shown in block-ing and co-operating scores The variance measure was calculated as followsMean values for activate periods and calm periods were calculated separatelyfor block and co-operate conditions Mean values were taken from each indi-vidual score in that condition and squared and summed S [mean value for con-dition - individual score]2 This gave a summation of variability around themean for each activate calm and rest period for block and cooperate condi-tions It was used in conjunction with the mean level of EDA activity givingan indication of the motion of the wave (ie the variability of the EDA activi-ty) rather than the overall level

VVIQ

Anecdotally the use of appropriate imagery (eg visualize being surroundedby a shielding white halo of white light) is suggested to be an effective blockingtechnique (eg Hope 1983) We therefore wished to see if participants who re-ported experiencing vivid imagery with ease are perhaps more effective block-ers than those who do not report clear mental imagery Descriptive data wasgathered to explore any trends in PSI scores for blockingco-operating forthose participants with VVIQ scores ranging from low (ie reporting vividmental imagery) to high (ie reporting little or no mental imagery)

Qualitative Reports

We also wanted to report data illustrating what kind of different strategiesparticipants used when they were blocking and co-operating These qualitativereports were gathered in the form of the participantsrsquo written responses toopen-ended questions on a post-session questionnaire Watt and Ravenscroftindependently coded these written responses into categories and cross-checked to ensure coding agreement

Sex and Experimenter Effects

We also examined the data for sex effects and experimenter effects becausethese variables could confound the interpretation of any overall effects Final-ly a post hoc analysis was conducted on influenceeinfluencer sex pairings Ina study on the unconscious detection of remote staring Schlitz and LaBerge(1997) found the greatest effect for opposite-sex observer-subject pairs

Study One Results and Discussion

Participants

Thirty-two individuals took part as influencees (12 females 20 males) theirages ranged from 19 to 63 years (mean age 285 years)

Incomplete Data

One participant declined to give their age on the PIF form so the descriptivestatistics on age do not include this participant (so N = 31) Also due to a tech-nical hitch the exploratory variance scores were lost for one participant so N = 31 for this measure There were no drop-outs from this study

Remote Influence Measure

Table 1 presents a summary of the remote influence data in the form ofmean Percentage Score Index for co-operating and blocking conditions Therewas no significant difference between PSI for co-operating and blocking con-ditions (t = - 202 df = 31) Therefore Hypothesis 1 (predicting significantlygreater PSI scores for co-operating condition compared to blocking condition)was not supported by these results As planned the PSI scores for co-operatingand blocking runs were combined to examine Hypothesis 2 (of an overall re-mote mental influence effect in line with the calming and activating intentionsof the influencee) Mean PSI scores were exactly at chance expectation (t = - 031 df = 31) therefore there was no support for Hypothesis 2

Exploratory Questions

Variance Measure

Table 2 presents summary data for the exploratory measure of the amount ofvariance shown in each condition (lt 50 represents reduced variance gt 50 rep-resents increased variance) There was no significant difference between vari-ance scores for co-operating and blocking conditions (t = 835 df = 30) andwhen the conditions were combined variance did not depart from chance ex-pectation (t = - 244 df = 30) It is interesting to note however that the vari-ance measure gives a much wider range of scores than the PSI measure

Blocking vs Co-operating 523

TABLE 1Summary Data Mean PSI Scores

(N = 32 MCE = 50 range in brackets)

Co-operate PSI Block PSI Combined PSI

4996 5003 5000(4692- 5215) (4653- 5404) (4853- 5309)

TABLE 2Summary Data Mean Variance Scores(N = 31 MCE = 50 range in brackets)

Co-operate Variance Block Variance Combined Variance

5089 4864 4977(2648- 7012) (2492- 6484) (4116- 6297)

524 C Watt et al

suggesting that the variance measure may potentially be a more sensitive indi-cator of remote influence effect than PSI

Vividness of Mental Imagery

Low VVIQ scores represent participants reporting highly vivid imagery andhigh VVIQ scores represent those reporting little or no imagery The meanscore was 4275 (range from 21 to 80) Using a Pearson correlation we foundthat there was no evidence to suggest that those participants with low VVIQscores (high imagery) were better blockers than those with high scores (VVIQvs PSI blocking r = - 113) We also found that there was no evidence of a cor-relation between vividness of imagery and overall remote influence effectsusing the combined Percentage Score Index measure (VVIQ vs PSI overall r =- 077)

Influenceesrsquo Blocking and Co-operating Strategies

Influencees completed a questionnaire (after the session was over) askingthem to describe what strategies they used for blocking and co-operating Cat-egorizing these responses we found a difference in the kinds of strategies usedwhen blocking compared to co-operating as shown in Table 3 In the blockingcondition the majority of participants (469) used visual imagery strategiesFewer (281) participants used a combination of imagery and cognitivestrategies (such as repeating words related to blocking) while 25 reportedusing purely cognitive strategies for blocking (such as adopting a resistant anduncooperative attitude or thinking of irrelevancies) In the co-operating con-dition fewer participants used visual imagery alone (333) The majority ofparticipants (424) reported that their strategies for co-operating includedadopting a feeling of passiveness openness or relaxation whereas no partici-pants adopted this kind of strategy for the blocking condition 121 of partic-ipants used cognitive strategies alone (such as adopting a co-operative atti-tude) and the same proportion used a combination of visual imagery andcognitive strategies

Therefore there was a tendency for participants to use visual imagery forblocking and relaxationpassivity or visual imagery for co-operating This wasdespite the fact that we asked participants to attempt to adopt similar levels ofactivation for both co-operating and blocking strategies There was also a ten-dency not surprisingly for those participants with low VVIQ scores (ie vivid

TABLE 3Influenceesrsquo Mental Strategies ()

Vis Im Cognitive VisCog OpenRelax

Blocking 469 250 281 00Co-operating 333 121 121 424

imagers) to use visual imagery strategies Those who had high VVIQ scoresalso used visual imagery but more often used this in combination with othertactics

Other Qualitative Reports from Participants According to their post-ses-sion questionnaire responses participants generally felt comfortable about theattempt to be influenced (N = 24) Three participants said they felt apprehen-sive or uncomfortable with the idea of being influenced Interestingly of thesethree two were actively involved with parapsychology and were aware of theharmful psi implications discussed at the beginning of this report The thirdwas a psychology student who had attended a lecture given by Watt in whichthe possibilities of harmful psi were discussed Another group of three partici-pants said they would be uncomfortable with the idea of being influenced ifthey believed such influence was possible

Seven participants said they found it was difficult to sustain concentrationfor the 10 minutes that comprised each condition Five participants found thescreensaver distracting or disturbing Two participants specifically mentionedthat they found blocking a more taxing task than co-operating

Sex Effects

There were more males (N = 20) than females (N = 12) in this study howev-er there was no significant difference in PSI scores for remote physiology in-fluence between males and females (male PSI vs female PSI t = 1085 male PSI = 4987 female PSI = 5021)

Experimenter Effects

We studied whether there was any difference in PSI scores obtained by thethree experimenters The results are shown in Table 4 There seems to be nosign of experimenter effects in study one

InfluenceeInfluencer Sex Pairings

Table 5 gives the mean PSI scores for different influenceeinfluencer pair-ings Taking the combined PSI scores it can be seen that the highest remote in-fluence effect is for FM pairs followed by MF pairs followed by FF pairsand the lowest scoring is for MM pairs Post hoc it was found that the mixedsex pairings showed significantly higher remote influence scores than the same

Blocking vs Co-operating 525

TABLE 4Remote Influence Results by Experimenter

Experimenter N sessions Coop PSI Block PSI Combined PSI

CW 12 4983 5011 4997ZMcD 10 5015 4998 5006JR 10 4994 4999 4996

526 C Watt et al

sex pairings (t = - 2016 df = 30 p = 053 2 - t) This replicates the effect foundby Schlitz and LaBerge (1997)

Study One Conclusions

In Study 1 we developed a methodology to examine intentional resistance toa remote mental influence attempt (Watt et al 1997) We contrasted influ-enceesrsquo physiological activity when they were mentally co-operating with aremote attempt to influence their physiology with physiological activity wheninfluencees were mentally blocking the influence attempt We found no evi-dence for any remote influence effect overall and no difference betweenblocking and co-operating conditions Our qualitative measures yielded someinteresting information about participantsrsquo mental strategies and reactions dur-ing the remote influence session We found no difference between experi-menters in remote influence scores but we did find that mixed sex influ-enceeinfluencer pairings scored more highly than same sex pairings

Study Two Method

Study 2 was planned to explore in more depth the nature of the relationshipbetween aspects of personality and cognition and psi performance Partici-pants took part in three testing sessions approximately one week apart Thefirst consisted of various psychological measures the second consisted of aforced choice ESP task and the third consisted of a replication of the blocking-co-operating study (Study 1) described above The psychological measureswere not scored until after completion of the remote influence session Partici-pants received feedback about their forced choice ESP performance at the con-clusion of the second session For ease of interpretation the present paper isrestricted to a description of the third testing session which was essentially areplication of study one with fewer influencers and greater statistical power

We introduced some slight changes to the methodology of Study 2 based onour experiences with Study 1 At the start of each of the two runs in Study 1the participant was given auditory instructions either to ldquoblock nowrdquo or ldquoco-operate nowrdquo We observed that the auditory instructions appeared to affect theparticipantrsquos EDA as if they were startled As planned the first recordingepoch in each run was not used in analysis to give the participantrsquos EDA time

TABLE 5Mean PSI Scores for Different InfluenceeInfluencer Sex Pairings

Pairing N Coop PSI Block PSI Combined PSI

M-F 6 4932 5069 5001F-M 6 5043 5034 5048F-F 6 5022 4952 4993M-M 14 4988 4975 4981

to settle down However in order to remove any remaining potential artifactu-al effects on EDA in Study 2 only visual instructions were given

Another difference in methodology concerned the screensaver viewed bythe participants during the experimental session In Study 1 some participantsreported that the series of patterns produced by the computer were being dis-played at too fast a rate such that some participants found it distracting or dis-turbing Again we felt that this may have an artifactual effect on EDA so wedecided for study two to reduce the speed of the screensaver

Thirdly in Study 2 we did not give participants detailed information abouttheir remote influence performance until the feedback letters that were distrib-uted at the conclusion of the entire study The reason for this change was mere-ly to streamline the procedure on what was already a time-consuming experi-ment Finally in another streamlining change we did not take VVIQ measuresin Study 2

Planned Hypotheses and Analyses

These were the same as with Study 1 That is hypothesis 1 predicted higherPSI scores for the co-operating condition than for the blocking condition(measured with a related t-test one-tailed) If no significant difference wasfound between blocking PSI and co-operating PSI scores the two would becombined to look at overall remote influence effect (hypothesis 2 single-meant-test one-tailed) If a significant difference was found with hypothesis 1 theanalysis for hypothesis 2 would be conducted with the data for co-operatingsessions only

Exploratory Questions

As with study one we took an exploratory variance measure of remote EDAinfluence We also gathered qualitative descriptions of the mental strategiesparticipants adopted whilst blocking and co-operating We also looked at sexeffects experimenter effects and post hoc influenceeinfluencer sex pairings

Participants

Study 2 had two experimenterinfluencers (Watt amp Ravenscroft) After aparticipant was recruited (usually by Ravenscroft) they were assigned to oneof the two experimenters for all three testing sessions This assignment wasdone unsystematically according to the preferences and availability of the ex-perimenters As planned 50 volunteer participants were drawn from the gener-al participant pool at the Koestler Parapsychology Unit and by word ofmouth All participants completed the Participant Information Form and aConsent form prior to the study The participants consisted of 33 females and17 males Average age was 41 years range from 22 years to 76 years of age Inorder to increase the studyrsquos relevance to the general population undergradu-ate psychology students were not recruited as participants The main criterion

Blocking vs Co-operating 527

528 C Watt et al

for participation was that the participants were interested in and open-mindedabout parapsychology Again there were no drop-outs from this study Allparticipants received a detailed personal feedback letter at the conclusion ofthe study

Study Two Results and Discussion

Remote Influence Measure

PSI scores are summarized in Table 6 which shows that the mean PSI scorewhen co-operating was nonsignificantly lower than mean PSI score whenblocking (t = - 595 df = 49) Hypothesis 1 was therefore not supported Asplanned in the absence of any significant difference between blocking and co-operating PSI scores the two conditions were combined to test hypothesis 2which predicted overall significant remote mental influence performance Thecombined PSI score was significantly greater than mean chance expectation of50 (t = 1806 df = 49 p = 038 1 - t) Therefore hypothesis 2 was supportedgiving evidence of an overall remote influence effect upon psychophysiologyin the prespecified direction

The failure to support hypothesis 1 confirms what we found in Study 1 thatthere was no significant difference between blocking and co-operating condi-tions on remote influence scores Indeed both studies found a slight effect inthe opposite direction to that predicted that is slightly higher PSI scores whenblocking was the influenceersquos intention than when co-operating was their in-tention It was encouraging to find significant support for a remote influenceeffect in the predicted direction for blocking and co-operating conditions com-bined No such trend was found in the previous study though with a smallernumber of participants (N = 32) the earlier study had lower statistical powerthan the present study

Exploratory Questions

Variance Measure

Table 7 shows the results for the exploratory measure of EDA variance lt 50represents reduced variance gt 50 represents increased variance We found nodifference between variance scores for cooperate versus block conditions(t = 150 df = 49) When these are combined the resulting figure does not de-viate from chance expectation (t = 440 df = 49) Thus as in Study 1 it seems

TABLE 6Summary Data Mean PSI Scores

(N = 50 MCE = 50 range in brackets)

Co-operate PSI Block PSI Combined PSI

5008 5019 5016(4821- 5221) (4807- 5557) (4900- 5296)

that the exploratory variance measure does not reveal any additional indica-tion of the operation of a remote mental influence in this study However thewider range of variance scoring continues to suggest that this measure mayhave greater potential sensitivity than the summary PSI score

Influenceesrsquo Blocking and Co-operating Strategies

We were interested in learning more about what particular strategies partici-pants chose to adopt when blocking and co-operating As in Study 1 a post-session questionnaire was used to gather this information Table 8 shows thepercentage of participants who adopted each kind of strategy

The table shows a very similar pattern to that found in Study 1 The mainpoint of interest is that almost no participants took a passive approach toblocking (ie adopting an openrelaxing strategy) A much higher proportionpreferred to adopt this approach when co-operating The preferred mode forboth conditions was to use some form of visual imagery such as imagining ashield or a screen when blocking and imagining an open channel such as ariver when co-operating

The majority of participants reported no concerns over the possibility ofhaving their physiology remotely influenced by another person as found instudy one

Sex Effects

As in study one the sexes were not equally represented among participantsWhereas Study 1 had an excess of male participants in Study 2 twice as manyfemales (N = 33) as males (N = 17) participated It was therefore important tolook for sex differences since any such differences could confound the inter-pretation of overall trends We found no difference between sexes when co-op-erating (female mean PSI score = 5008 male mean PSI score = 5006 t = 107 df = 48) There was a nonsignificant trend for females to have higher

Blocking vs Co-operating 529

TABLE 7Summary Data Mean Variance Scores(N = 50 MCE = 50 range in brackets)

Co-operate Variance Block Variance Combined Variance

5076 5054 5034(3164- 7279) (3572- 6813) (3749- 6671)

TABLE 8Influenceesrsquo Mental Strategies ()

Visual Imagery Cognitive VisCog OpenRelax

Blocking 64 26 6 2Co-operating 60 8 12 20

530 C Watt et al

PSI scores than males when blocking (mean PSI scores 5030 and 4998 re-spectively t = 992 df = 48) Combining blocking and co-operating conditionsthere was no overall difference between the sexes (mean PSI scores 5020 forfemales 5008 for males t = 665 df = 48)

Experimenter Effects

Study 2 found a statistically significant overall remote influence effect(combined PSI = 50155 t = 1806) Table 9 shows the breakdown of remoteinfluence scores for the two experimenters CW and JR Looking at the overalldata for the two experimenters we see that CWrsquos sessions were associatedwith most of the remote influence effect However due to the smaller N CWrsquossessions are not independently significant (t = 1632) and there is not a signifi-cant difference between CW and JRrsquos overall PSI scores (t = 678) Post hocwe calculated effect size r according to the formula r = [t2(t2+df)]12 for the twoexperimenters For JR (t = 817 df = 22) we found a small effect sizer = +17 For CW (t = 1632 df = 26) the effect size was moderate r = +30This may be some post hoc indication of an experimenter effect in Study 2

InfluenceeInfluencer Sex Pairings

As can be seen from Table 10 Study 2 found exactly the opposite trend tothat found in Study 1 On a post hoc test same sex pairings scored significant-ly higher on the remote influence PSI measure than mixed sex pairings(t = - 2034 df = 48 p = 048 2 - t) This contradicts Schlitz and LaBergersquos(1997) finding of greater autonomic detection of remote staring for mixed sexpairs

TABLE 9Remote Influence Results by Experimenter

Experimenter N sessions Coop PSI Block PSI Combined PSI

CW 27 5011 5026 5021JR 23 5003 5010 5009

TABLE 10Mean PSI Scores for Different InfluenceeInfluencer Sex Pairings

Pairing N Coop PSI Block PSI Combined PSI

MF 11 4977 5009 4997FM 17 4983 5021 5003FF 16 5035 5038 5038MM 6 5060 4978 5027

General Summary and Conclusions

We have described the findings of two studies exploring the limits of remotemental influence effects We wished to see whether the mental strategiesadopted by the influencee would affect remote influence scores In particularcould an influencee block or resist a remote influence attempt We recorded in-fluenceesrsquo electrodermal activity while the influencer attempted to calm to ac-tivate or to have no influence over the influenceesrsquo physiology In each studythere were two conditions reflecting the influenceesrsquo designated mental strate-gies in response to the remote influence attempt blocking and co-operating

There were only two formal hypotheses 1) that remote influence scoreswould be significantly higher during the co-operating condition than duringthe blocking condition and 2) that overall there would be significant evidenceof remote mental influence on physiology in the designated direction (iecalming when calming was the influencerrsquos intention activating when activat-ing was the influencerrsquos intention) Contrary to prediction Study 1 found thatremote influence effects were higher in the block condition than in the co-op-erate condition (but not to a significant extent) Also overall remote influenceperformance was at chance level This study therefore found no support for ei-ther of the formal hypotheses Study 2 consisted of a replication of Study 1with a few minor adjustments plus additional measures that were not reportedin this paper Again contrary to prediction remote influence scores were non-significantly higher in the block condition than in the co-operate conditionThis replicates the trend found in Study 1 failing to support hypothesis oneThese results suggest that the mental strategy adopted by the influencee in theface of a remote influence attempt has no effect on remote influence perfor-mance Referring back to the questions posed in the introduction then wehave not found any evidence to suggest that it is possible to resist a remote in-fluence effect and there is as yet no indication of the possible limiting condi-tions for remote mental influence

Unlike Study 1 Study 2 found statistically significant evidence of an overallremote influence effect in the prespecified direction (ie calming of EDAwhen that was the influencerrsquos intention activation of EDA when that was theinfluencerrsquos intention) In order to facilitate comparison with a recent meta-an-alytic review of remote influence studies we calculated the effect size (r) ac-cording to the formula r = [t2(t2+df)]12 For their review of 19 EDA influencestudies Schlitz and Braud (1997) found an average effect size of r = +25 OurStudy 1 with overall t = - 031 df = 31 gives a negligible effect size of r = - 006 Study 2 with overall t = 1806 df = 49 gives r = +25 Study 2 there-fore replicates the mean effect size found in EDA studies conducted by otherresearchers in other parapsychology laboratories Post hoc we can combinethe data of the two studies for greater statistical power Overall t = 1176 df = 81 giving a small effect size of r = +13

We can find no clear reason why Study 2 found an EDA influence effectwhile Study 1 did not A number of factors are worth considering 1) The

Blocking vs Co-operating 531

532 C Watt et al

significant result could be due to chance or undetected artifact 2) Study 2 im-plemented what we considered to be minor methodological improvements thatmight have removed some artifactual fluctuations from the EDA data that weobserved in study one 3) Each experimenter in Study 2 conducted three testingsessions with their participant By the third session in which the remote influ-ence measure was taken experimenters were on quite friendly terms with theirparticipants perhaps more so than in Study 1 where each participant came infor only one testing session 4) The participants in Study 2 were more represen-tative of the general population we felt than those in Study 1 Study 1 in-volved more undergraduate psychology students and associates of the experi-menters 5) The experimenters in study one were less experienced with theremote influence protocol than when they conducted Study 2

In addition to the formal hypotheses we reported a number of exploratorymeasures Study 1 found that influenceesrsquo self-reported vividness of visual im-agery appeared to be unrelated to the remote influence performance In bothstudies influenceesrsquo qualitative descriptions of the actual mental strategiesthey employed when blocking and co-operating revealed a tendency for differ-ent strategies to be preferred for the different conditions The use of visual im-agery was popular for both blocking (eg visualizing being surrounded by aprotective cocoon) and co-operating (eg visualizing a river flowing betweeninfluencer and influencee) However a considerable proportion of participantsreported adopting an open or relaxed frame of mind when co-operating whilstalmost no participants chose this strategy when blocking This was despite thefact that prior to each session we encouraged participants to adopt comparablestrategies for blocking and co-operating

It was interesting and perhaps surprising to note that in both studies themajority of participants reported no feelings of concern or discomfort at thepossibility of having their physiology remotely influenced Those who wereconcerned tended to be those who knew more about parapsychology and thepossible harmful applications of remote influence A small proportion of par-ticipants was unconcerned because they did not believe such an influence waspossible in the first place

Both studies employed an exploratory variance measure of EDA influenceThis measure showed a greater range of scoring overall thus perhaps greaterpotential sensitivity to a remote influence effect However the variance mea-sure showed no indication of a remote influence effect To date researchershave employed a variety of indicators of remote influence as the literaturegrows progress may be made on identifying which is the most appropriate in-dicator

Neither study found significant sex differences in remote influence effectsthough in both studies female influencees tended to have higher remote influ-ence scores than male influencees

In terms of their overall remote influence effects there appeared to be nodifference between experimenterinfluencers in Study 1 (r = +06 - 05 and

- 06) In Study 2 effect size measures indicated that the experimenters dif-fered in remote influence effect sizes (r = +17 and +30) Due to the smaller Nthis difference was not found to be statistically significant Post hoc combin-ing Studies 1 and 2 for experimenters JW and CW (ZMcD was not an experi-menter in Study 2) the effect sizes are r = +08 and r = +20 respectively

We conducted a post hoc analysis of the effects of influenceeinfluencer sexpairings on remote influence scores Our two studies came up with contradic-tory findings each statistically significant Study one found that mixed sexpairings had significantly higher remote influence scores than same sex pair-ings This replicated the post hoc finding of Schlitz and LaBerge (1997) in con-nection with the autonomic detection of remote observation Study 2 foundthat same sex pairings had significantly higher remote influence scores thanmixed sex pairings Our contradictory findings do not help to clarify the pic-ture on the importance or otherwise of sex pairings It is quite possible that sexis not the most salient aspect of the dynamic between pairs Age personalityand familiarity are likely to be important variables moderating the relationshipbetween influencer and influencee

There are many more variables that we could look at but as the number ofanalyses multiplies so does the opportunity for spurious significant resultsFor this reason we have named only two formal hypotheses for each experi-ment All other analyses are regarded as exploratory and any trends that wehave highlighted above would require replication

In conclusion these two studies have found no evidence of a difference inremote influence effects when influencees are asked to adopt cooperative men-tal strategies compared to when they are asked to adopt blocking mental strate-gies In fact contrary to expectation both studies found slightly higher remoteinfluence effects for the blocking condition compared to the co-operating con-dition On this basis we have found no indication of limiting conditions for re-mote influence effects Therefore while anecdotal advice exists on how to de-fend oneself against a remote physiological influence attempt in the realworld laboratory evidence to date has provided no support for this advice

Given the importance of this question both in practical terms and in theoret-ical terms we argue that this line of research deserves continued investigationIf a way could be found to conduct research with a more realistic scenario mdash where the remote influence was not just neutrally and abstractlydefined as unwelcome mdash then our research might have more meaningful thingsto say about the parameters of remote influence It is interesting and perhapssurprising to note that when Watt has lectured on this topic to informed lay au-diences the feedback has often been that researchers must use more potential-ly harmful or disruptive remote influence (of course with the informed consentof the participants) While there are clear ethical inhibitions when consideringpossible physiological influence perhaps research using higher level influencewould be more acceptable For example one could attempt to disrupt or en-hance the concentration of chess players who were aware of the influence

Blocking vs Co-operating 533

534 C Watt et al

attempt and who were concurrently blocking or co-operating with the influ-ence attempt Research already exists to suggest that remote influence does notjust operate at physiological levels Attention focusing for example can be fa-cilitated by a remote helper (Brady amp Morris 1997 Braud et al 1995) Futureresearch in this area must therefore expand beyond physiological influences

Acknowledgements

The authors gratefully acknowledge the financial support they receivedfrom the Institut fuumlr Grenzgebiete der Psychologie und PsychohygieneFreiburg Germany while conducting these studies We would also like tothank Dr Paul Stevens for generously helping to solve software difficultieswhen preparing for these studies and the anonymous reviewer for helpfulcomments Address reprint requests to Dr C Watt Department of PsychologyUniversity of Edinburgh 7 George Square Edinburgh EH8 9JZ Scotland UK

ReferencesBrady C amp Morris R L (1997) Attention focusing facilitated through remote mental interac-

tion A replication and exploration of parameters In R Wiseman (Ed) Proceedings of Pre-sented Papers The Parapsychological Association 40th Annual Convention held in conjunc-tion with The Society for Psychical Research pp 73ndash91

Braud W G (1985) Blockingshielding psychic functioning through psychological and psychictechniques A report of three preliminary studies In RA White and J Solfvin (Eds) Re-search in Parapsychology 1984 Metuchen NJ Scarecrow Press pp 42ndash44

Braud W G amp Schlitz M J (1991) Consciousness interactions with remote biological systemsAnomalous intentionality effects Subtle Energies 2 1

Braud W G Schlitz M J Collins J and Klitch H (1985) Further studies of the bio-PK effectFeedback blocking specificitygenerality In RA White amp J Solfvin (Eds) Research inParapsychology 1984 Metuchen NJ Scarecrow Press pp 45ndash48

Braud W G Shafer D McNeill K and Guerra V (1995) Attention focusing facilitatedthrough remote mental interaction Journal of the American Society for Psychical Research89 103

Cohen J (1997) Statistical Power Analysis for the Behavioral Sciences New York AcademicPress

Dalton K (1994) Remotely influenced ESP performance in a computer task A preliminarystudy Proceedings of the 37th Annual Convention of the Parapsychological Association pp95ndash103

Dossey L (1994) Healing and the mind Is there a dark side Journal of Scientific Exploration 873

Halifax-Grof J (1974) Hex death In A Angoff amp D Barth (Eds) Parapsychology and Anthro-pology Proceedings of an International Conference Held in London England August 29ndash311973 NY Parapsychology Foundation

Hope M (1983) Practical Techniques of Psychic Self-Defence Wellingborough Northampton-shire Aquarian Press

Marks DF (1973) Visual imagery differences in the recall of pictures British Journal of Psy-chology 64 17

Nash CB (1982) Psychokinetic control of bacterial growth Journal of the Society for PsychicalResearch 51 217

Radin DI Taylor RK and Braud WG (1995) Remote mental influence of human electroder-mal activity A pilot replication European Journal of Parapsychology 11 19

Schlitz M amp Braud W G (1997) Distant intentionality and healing Assessing the evidence Al-ternative Therapies 3 62

Schlitz M J amp LaBerge S (1997) Covert observation increases skin conductance in subjectsunaware of when they are being observed A replication Journal of Parapsychology 61 185

Schmeidler G R (1958) Agent-percipient relationships Journal of the American Society forPsychical Research 52 47

Schmeidler G R (1961) Evidence for two kinds of telepathy International Journal of Parapsy-chology 3 5

Tart C T (1979) A survey of expert opinion on potentially negative uses of psi United Statesgovernment interest in psi and the level of research funding of the field In WG Roll (Ed) Re-search in Parapsychology 1978 Metuchen NJ Scarecrow Press pp 54

Tart C T amp Labore C M (1986) Attitudes towards strongly functioning psi A preliminary sur-vey Journal of the America Society for Psychical Research 80 163

Tart C T (1986) Psychicsrsquo fears of psychic powers Journal of the American Society for Psychi-cal Research 80 278

Watt C A McDermott Z amp Ravenscroft J (1997) Exploring the limits of remote mental inter-action A comparison of blocking versus co-operating strategies In R Wiseman (Ed) Pro-ceedings of the 40th Annual Convention of the Parapsychological Association held in conjunc-tion with the Society for Psychical Research Brighton England pp 464ndash476

Blocking vs Co-operating 535

518 C Watt et al

Design

We used a within-subjects design (whereas Braud and Schlitz used a between-subjects design with different groups of volunteers blocking or co-operating)The independent variable was the influenceersquos strategy (blocking vs co-oper-ating) while the influencer attempted either to calm activate or have no influ-ence over the influenceersquos physiology The dependent variable was the influ-enceersquos physiology EDA being the principal measure but skin temperatureand finger pulse volume being recorded for exploratory purposes (we do notreport these exploratory measures in this paper) Our primary aim was to deter-mine whether there was a smaller remote influence effect when the influenceewas blocking compared to when they were co-operating with the influence at-tempt Our secondary aim was to ascertain whether there was any evidence ofan overall remote influence effect

Influence Schedule

The influence schedule that was followed by the influencer was composedof blocks of four epochs mdash either Rest-Activate-Rest-Calm (RARC) or Rest-Calm-Rest-Activate (RCRA) The ordering of each of these blocks of four wasrandomly determined by the computer as the session proceeded so that in totalthere was an equal number of activate and calm epochs Note that the randomsequence also serves to counter-balance the activatecalm conditions acrossthe duration of the study Each epoch was 20 seconds in duration The experi-mental session was thus split into two runs of equal length each having 33recording epochs ie eight blocks of four (RARC or RCRA) plus an extrarest at the end of each run There was a total of 66 recording epochs over theentire session Between the two runs there was a 150 second rest period duringwhich the influencee and the influencer remained in their rooms

The duration of each epoch and the total number of epochs were chosen inorder to maximize the trade-off between gathering sufficient data and keepingthe duration of the session (2 runs of 10 min) at a comfortable length for par-ticipants Additionally we felt that 20 second epochs would be sufficient toenable any stimulus-response effect to manifest in EDA A liberal strategy(that used by Radin et al 1995) was adopted to specify the criteria for includ-ing data if a session was aborted due to technical faults with the psychophysi-ology apparatus data was included from any session that had at least one calmand one activate period (ie a total of four recording epochs)

Data Treatment

Braudrsquos Percentage Score Index (PSI as described in Braud amp Schlitz1991) per session was the unit of analysis used in this study PSI is a conserva-tive summary of the results of an entire session and is calculated by dividingthe sum of EDA activity over all Activate epochs by the sum of all calm andactivate EDA [PSI = S A( S C+ S A)] The rest epochs were not used in this

analysis Their function was primarily to separate the calm and activateepochs from one another and to separate any lingering influence that may havebeen left at the end of the influencerrsquos calm or activate epochs In the absenceof any remote influence effect PSI equals 50 A PSI of greater than 50 rep-resents a remote influence effect in the predicted direction (ie calming whenthat was the influencerrsquos intention activating when that was the influencerrsquosintention) A PSI of less than 50 represents a remote influence effect in thedirection opposite to that predicted For the present study we calculated PSIseparately for each participantrsquos blocking and co-operating condition plus wecalculated an overall PSI score for each participant

Participants

The influencees were volunteer participants drawn from the authorsrsquo con-tacts in the Department of Psychology at Edinburgh University and includingundergraduate psychology students recruited by e-mail and by word of mouthThere were three experimentersinfluencers overall (Watt Ravenscroft andMcDermott) but in each testing session there were only two individuals mdash thevolunteer influencee and the influencer Each experimenter also took part asinfluencer in one session

Written Materials

Each volunteer was given a one page information sheet Once they had readthis they were given a 34-item participant information form (PIF) and a 16-item Vividness of Visual Imagery Questionnaire (Marks 1973) After dis-cussing the procedure with the influencer each participant was asked to sign aconsent form that assured confidentiality and noted their right to withdrawfrom the study at any time Following completion of the DMILS procedure in-fluencee and influencer both completed one-page qualitative post-sessionquestionnaires in which they were invited to describe the kinds of strategiesthat they used during the testing session and to give their subjective impres-sions of the session A letter of thanks and a brief overview of the results weregiven to participants at the end of the study

Technical Apparatus

Two rooms were used for the testing sessions The influenceersquos room wassound-attenuated with double-doors and was located approximately 20 metersaway from the influencerrsquos room down a short flight of stairs The DMILS ap-paratus (an I-410 multichannel physiological monitoring unit from J amp J Sys-tems with source code from Physiodata) was set up between these two roomsand was run on two time-synchronized Pentium computers In the influenceersquosroom a monitor displayed written instructions and a mildly engaging visualdisplay (a screen saver shown during the actual data collection epochs) to theinfluencee This room also contained speakers that gave auditory instructions

Blocking vs Co-operating 519

520 C Watt et al

(ldquohey block nowrdquo ldquohey co-operate nowrdquo and ldquosession over thank yourdquo) tothe influencee just prior to each of the two 10-minute runs duplicating thewritten instructions that appeared on the computer monitor The physiologysensors that were attached to the influenceersquos fingers were located in the influ-enceersquos room The signals from the sensors were relayed along a cable andprocessed by a computer located in the influencerrsquos room Therefore the com-puter in the influenceersquos room was merely used to provide instructions and avisual display to the influencee The computer in the influencerrsquos room con-trolled the DMILS software and processes and stored the psychophysiologydata

Procedure

The influencer and participant initially sat down in the reception room in theparapsychology suite for a further description of procedure and returned theircompleted PIFimagery questionnaire to the influenceer who went over it withthe participant Any questions that were raised by the participant prior to test-ing were answered here (This question-and-answer session usually lasted fif-teen minutes) The participant then gave their informed consent to the studyThe influencer also suggested various mental approaches the participant mightadopt for the blocking and co-operating conditions (some examples weregiven in the information sheet) These suggestions included both visual im-agery (eg visualizing a shielding cocoon in the blocking condition) and cog-nitive strategies (eg adopting a stubborn and uncooperative frame of mind inthe blocking condition) Participants were informed that different strategiesmight suit different people and they were to do whatever felt natural for themThe influencer then took the participant to the testing room and demonstratedthe auditory instructions so that the participant would not be startled by thesewhen they occurred during the session itself The participant was also taken tosee where the influencer was sitting and the influencer described what theywould be doing when they were attempting to activate or calm the participant

Once the preliminaries were finished the influencee was seated in a partiallyreclined chair in front of the computer monitor and the physiology sensorswere attached to their fingers by the influencer The room was softly illuminat-ed by a desk lamp with an orangered filter At this stage the influencer checkedthat the physiology signals were being transmitted to the computer in theirroom Returning to the influencee the influencer set a start time on the influ-enceersquos computer This was on average two to three minutes later The influ-encer then returned to their room closing all doors and waited for the start ofthe session

Physiology Data Collection As the session began the influencee was in-structed visually and auditorially either to ldquohey block nowrdquo or ldquohey co-oper-ate nowrdquo The influenceersquos monitor then displayed a colorful and dynamicscreen-saver that was intended to be engaging without being particularlyarousing or relaxing At the same time the influencer was following the influ-

ence schedule as instructed by the computer observing the real-time feedbackdisplay of the participantrsquos EDA when desired At the end of the first run whichlasted approximately ten minutes the influencer saw a 150 second count downon the computer before the second run began At the same time the influenceesaw a message about the strategy to adopt for the forthcoming second run (ieto block or to co-operate whichever strategy was not used for the first run) Atthe beginning of the second run the influencee was also instructed auditoriallyeither to ldquohey block nowrdquo or ldquohey co-operate nowrdquo The dynamic visual dis-play resumed while the influencer followed a random influence schedule forapproximately ten minutes The entire session took about 25 minutes to com-plete At the end of the session the influencee was auditorially and visually in-formed ldquosession over thank yourdquo and waited until the influencer returned tothe influenceersquos room

Post-Testing Procedures The influenceersquos physiology sensors were re-moved by the influencer and both returned to the reception room to completethe post-session questionnaires Then the influencee was taken to the influ-encerrsquos room to see feedback on the results of their testing session There wasan option for some final discussion about the session after which the volun-teer was thanked for their participation and left Once all the data was collect-ed feedback on the results of the study as a whole was sent to participants

Planned Hypotheses and Analyses

Our main aim was to examine whether there was any evidence for blockingof a remote mental influence on physiology [EDA is the principal physiologymeasure here with per-session PSI scores representing the measure of theDMILS effect on EDA] H1 therefore predicted significantly greater PSIscores in the co-operating condition than in the blocking condition (using a re-lated t-test one-tailed) Our second question was whether there was any over-all evidence of remote mental influence on EDA This analysis would be con-ducted on the data of blocking and co-operating runs combined if nosignificant difference was found between these two conditions (as tested byH1) If H1 was supported H2 would be tested on the data of co-operating runsonly predicting PSI scores significantly greater than chance expectation(using a single-mean t-test of the PSI score vs MCE one-tailed)

Planned alpha level was 05 There was no correction for multiple analysisThis was dealt with by restricting the number of formal hypotheses (with one-tailed significance tests) to two in each study All other analyses were regardedas exploratory and two-tailed significance tests were used

Exploratory Measures

Variance

DMILS researchers do not yet appear to have identified what might be themost appropriate indicator of a remote influence effect Therefore in addition

Blocking vs Co-operating 521

522 C Watt et al

to the PSI score we looked at the amount of variance that was shown in block-ing and co-operating scores The variance measure was calculated as followsMean values for activate periods and calm periods were calculated separatelyfor block and co-operate conditions Mean values were taken from each indi-vidual score in that condition and squared and summed S [mean value for con-dition - individual score]2 This gave a summation of variability around themean for each activate calm and rest period for block and cooperate condi-tions It was used in conjunction with the mean level of EDA activity givingan indication of the motion of the wave (ie the variability of the EDA activi-ty) rather than the overall level

VVIQ

Anecdotally the use of appropriate imagery (eg visualize being surroundedby a shielding white halo of white light) is suggested to be an effective blockingtechnique (eg Hope 1983) We therefore wished to see if participants who re-ported experiencing vivid imagery with ease are perhaps more effective block-ers than those who do not report clear mental imagery Descriptive data wasgathered to explore any trends in PSI scores for blockingco-operating forthose participants with VVIQ scores ranging from low (ie reporting vividmental imagery) to high (ie reporting little or no mental imagery)

Qualitative Reports

We also wanted to report data illustrating what kind of different strategiesparticipants used when they were blocking and co-operating These qualitativereports were gathered in the form of the participantsrsquo written responses toopen-ended questions on a post-session questionnaire Watt and Ravenscroftindependently coded these written responses into categories and cross-checked to ensure coding agreement

Sex and Experimenter Effects

We also examined the data for sex effects and experimenter effects becausethese variables could confound the interpretation of any overall effects Final-ly a post hoc analysis was conducted on influenceeinfluencer sex pairings Ina study on the unconscious detection of remote staring Schlitz and LaBerge(1997) found the greatest effect for opposite-sex observer-subject pairs

Study One Results and Discussion

Participants

Thirty-two individuals took part as influencees (12 females 20 males) theirages ranged from 19 to 63 years (mean age 285 years)

Incomplete Data

One participant declined to give their age on the PIF form so the descriptivestatistics on age do not include this participant (so N = 31) Also due to a tech-nical hitch the exploratory variance scores were lost for one participant so N = 31 for this measure There were no drop-outs from this study

Remote Influence Measure

Table 1 presents a summary of the remote influence data in the form ofmean Percentage Score Index for co-operating and blocking conditions Therewas no significant difference between PSI for co-operating and blocking con-ditions (t = - 202 df = 31) Therefore Hypothesis 1 (predicting significantlygreater PSI scores for co-operating condition compared to blocking condition)was not supported by these results As planned the PSI scores for co-operatingand blocking runs were combined to examine Hypothesis 2 (of an overall re-mote mental influence effect in line with the calming and activating intentionsof the influencee) Mean PSI scores were exactly at chance expectation (t = - 031 df = 31) therefore there was no support for Hypothesis 2

Exploratory Questions

Variance Measure

Table 2 presents summary data for the exploratory measure of the amount ofvariance shown in each condition (lt 50 represents reduced variance gt 50 rep-resents increased variance) There was no significant difference between vari-ance scores for co-operating and blocking conditions (t = 835 df = 30) andwhen the conditions were combined variance did not depart from chance ex-pectation (t = - 244 df = 30) It is interesting to note however that the vari-ance measure gives a much wider range of scores than the PSI measure

Blocking vs Co-operating 523

TABLE 1Summary Data Mean PSI Scores

(N = 32 MCE = 50 range in brackets)

Co-operate PSI Block PSI Combined PSI

4996 5003 5000(4692- 5215) (4653- 5404) (4853- 5309)

TABLE 2Summary Data Mean Variance Scores(N = 31 MCE = 50 range in brackets)

Co-operate Variance Block Variance Combined Variance

5089 4864 4977(2648- 7012) (2492- 6484) (4116- 6297)

524 C Watt et al

suggesting that the variance measure may potentially be a more sensitive indi-cator of remote influence effect than PSI

Vividness of Mental Imagery

Low VVIQ scores represent participants reporting highly vivid imagery andhigh VVIQ scores represent those reporting little or no imagery The meanscore was 4275 (range from 21 to 80) Using a Pearson correlation we foundthat there was no evidence to suggest that those participants with low VVIQscores (high imagery) were better blockers than those with high scores (VVIQvs PSI blocking r = - 113) We also found that there was no evidence of a cor-relation between vividness of imagery and overall remote influence effectsusing the combined Percentage Score Index measure (VVIQ vs PSI overall r =- 077)

Influenceesrsquo Blocking and Co-operating Strategies

Influencees completed a questionnaire (after the session was over) askingthem to describe what strategies they used for blocking and co-operating Cat-egorizing these responses we found a difference in the kinds of strategies usedwhen blocking compared to co-operating as shown in Table 3 In the blockingcondition the majority of participants (469) used visual imagery strategiesFewer (281) participants used a combination of imagery and cognitivestrategies (such as repeating words related to blocking) while 25 reportedusing purely cognitive strategies for blocking (such as adopting a resistant anduncooperative attitude or thinking of irrelevancies) In the co-operating con-dition fewer participants used visual imagery alone (333) The majority ofparticipants (424) reported that their strategies for co-operating includedadopting a feeling of passiveness openness or relaxation whereas no partici-pants adopted this kind of strategy for the blocking condition 121 of partic-ipants used cognitive strategies alone (such as adopting a co-operative atti-tude) and the same proportion used a combination of visual imagery andcognitive strategies

Therefore there was a tendency for participants to use visual imagery forblocking and relaxationpassivity or visual imagery for co-operating This wasdespite the fact that we asked participants to attempt to adopt similar levels ofactivation for both co-operating and blocking strategies There was also a ten-dency not surprisingly for those participants with low VVIQ scores (ie vivid

TABLE 3Influenceesrsquo Mental Strategies ()

Vis Im Cognitive VisCog OpenRelax

Blocking 469 250 281 00Co-operating 333 121 121 424

imagers) to use visual imagery strategies Those who had high VVIQ scoresalso used visual imagery but more often used this in combination with othertactics

Other Qualitative Reports from Participants According to their post-ses-sion questionnaire responses participants generally felt comfortable about theattempt to be influenced (N = 24) Three participants said they felt apprehen-sive or uncomfortable with the idea of being influenced Interestingly of thesethree two were actively involved with parapsychology and were aware of theharmful psi implications discussed at the beginning of this report The thirdwas a psychology student who had attended a lecture given by Watt in whichthe possibilities of harmful psi were discussed Another group of three partici-pants said they would be uncomfortable with the idea of being influenced ifthey believed such influence was possible

Seven participants said they found it was difficult to sustain concentrationfor the 10 minutes that comprised each condition Five participants found thescreensaver distracting or disturbing Two participants specifically mentionedthat they found blocking a more taxing task than co-operating

Sex Effects

There were more males (N = 20) than females (N = 12) in this study howev-er there was no significant difference in PSI scores for remote physiology in-fluence between males and females (male PSI vs female PSI t = 1085 male PSI = 4987 female PSI = 5021)

Experimenter Effects

We studied whether there was any difference in PSI scores obtained by thethree experimenters The results are shown in Table 4 There seems to be nosign of experimenter effects in study one

InfluenceeInfluencer Sex Pairings

Table 5 gives the mean PSI scores for different influenceeinfluencer pair-ings Taking the combined PSI scores it can be seen that the highest remote in-fluence effect is for FM pairs followed by MF pairs followed by FF pairsand the lowest scoring is for MM pairs Post hoc it was found that the mixedsex pairings showed significantly higher remote influence scores than the same

Blocking vs Co-operating 525

TABLE 4Remote Influence Results by Experimenter

Experimenter N sessions Coop PSI Block PSI Combined PSI

CW 12 4983 5011 4997ZMcD 10 5015 4998 5006JR 10 4994 4999 4996

526 C Watt et al

sex pairings (t = - 2016 df = 30 p = 053 2 - t) This replicates the effect foundby Schlitz and LaBerge (1997)

Study One Conclusions

In Study 1 we developed a methodology to examine intentional resistance toa remote mental influence attempt (Watt et al 1997) We contrasted influ-enceesrsquo physiological activity when they were mentally co-operating with aremote attempt to influence their physiology with physiological activity wheninfluencees were mentally blocking the influence attempt We found no evi-dence for any remote influence effect overall and no difference betweenblocking and co-operating conditions Our qualitative measures yielded someinteresting information about participantsrsquo mental strategies and reactions dur-ing the remote influence session We found no difference between experi-menters in remote influence scores but we did find that mixed sex influ-enceeinfluencer pairings scored more highly than same sex pairings

Study Two Method

Study 2 was planned to explore in more depth the nature of the relationshipbetween aspects of personality and cognition and psi performance Partici-pants took part in three testing sessions approximately one week apart Thefirst consisted of various psychological measures the second consisted of aforced choice ESP task and the third consisted of a replication of the blocking-co-operating study (Study 1) described above The psychological measureswere not scored until after completion of the remote influence session Partici-pants received feedback about their forced choice ESP performance at the con-clusion of the second session For ease of interpretation the present paper isrestricted to a description of the third testing session which was essentially areplication of study one with fewer influencers and greater statistical power

We introduced some slight changes to the methodology of Study 2 based onour experiences with Study 1 At the start of each of the two runs in Study 1the participant was given auditory instructions either to ldquoblock nowrdquo or ldquoco-operate nowrdquo We observed that the auditory instructions appeared to affect theparticipantrsquos EDA as if they were startled As planned the first recordingepoch in each run was not used in analysis to give the participantrsquos EDA time

TABLE 5Mean PSI Scores for Different InfluenceeInfluencer Sex Pairings

Pairing N Coop PSI Block PSI Combined PSI

M-F 6 4932 5069 5001F-M 6 5043 5034 5048F-F 6 5022 4952 4993M-M 14 4988 4975 4981

to settle down However in order to remove any remaining potential artifactu-al effects on EDA in Study 2 only visual instructions were given

Another difference in methodology concerned the screensaver viewed bythe participants during the experimental session In Study 1 some participantsreported that the series of patterns produced by the computer were being dis-played at too fast a rate such that some participants found it distracting or dis-turbing Again we felt that this may have an artifactual effect on EDA so wedecided for study two to reduce the speed of the screensaver

Thirdly in Study 2 we did not give participants detailed information abouttheir remote influence performance until the feedback letters that were distrib-uted at the conclusion of the entire study The reason for this change was mere-ly to streamline the procedure on what was already a time-consuming experi-ment Finally in another streamlining change we did not take VVIQ measuresin Study 2

Planned Hypotheses and Analyses

These were the same as with Study 1 That is hypothesis 1 predicted higherPSI scores for the co-operating condition than for the blocking condition(measured with a related t-test one-tailed) If no significant difference wasfound between blocking PSI and co-operating PSI scores the two would becombined to look at overall remote influence effect (hypothesis 2 single-meant-test one-tailed) If a significant difference was found with hypothesis 1 theanalysis for hypothesis 2 would be conducted with the data for co-operatingsessions only

Exploratory Questions

As with study one we took an exploratory variance measure of remote EDAinfluence We also gathered qualitative descriptions of the mental strategiesparticipants adopted whilst blocking and co-operating We also looked at sexeffects experimenter effects and post hoc influenceeinfluencer sex pairings

Participants

Study 2 had two experimenterinfluencers (Watt amp Ravenscroft) After aparticipant was recruited (usually by Ravenscroft) they were assigned to oneof the two experimenters for all three testing sessions This assignment wasdone unsystematically according to the preferences and availability of the ex-perimenters As planned 50 volunteer participants were drawn from the gener-al participant pool at the Koestler Parapsychology Unit and by word ofmouth All participants completed the Participant Information Form and aConsent form prior to the study The participants consisted of 33 females and17 males Average age was 41 years range from 22 years to 76 years of age Inorder to increase the studyrsquos relevance to the general population undergradu-ate psychology students were not recruited as participants The main criterion

Blocking vs Co-operating 527

528 C Watt et al

for participation was that the participants were interested in and open-mindedabout parapsychology Again there were no drop-outs from this study Allparticipants received a detailed personal feedback letter at the conclusion ofthe study

Study Two Results and Discussion

Remote Influence Measure

PSI scores are summarized in Table 6 which shows that the mean PSI scorewhen co-operating was nonsignificantly lower than mean PSI score whenblocking (t = - 595 df = 49) Hypothesis 1 was therefore not supported Asplanned in the absence of any significant difference between blocking and co-operating PSI scores the two conditions were combined to test hypothesis 2which predicted overall significant remote mental influence performance Thecombined PSI score was significantly greater than mean chance expectation of50 (t = 1806 df = 49 p = 038 1 - t) Therefore hypothesis 2 was supportedgiving evidence of an overall remote influence effect upon psychophysiologyin the prespecified direction

The failure to support hypothesis 1 confirms what we found in Study 1 thatthere was no significant difference between blocking and co-operating condi-tions on remote influence scores Indeed both studies found a slight effect inthe opposite direction to that predicted that is slightly higher PSI scores whenblocking was the influenceersquos intention than when co-operating was their in-tention It was encouraging to find significant support for a remote influenceeffect in the predicted direction for blocking and co-operating conditions com-bined No such trend was found in the previous study though with a smallernumber of participants (N = 32) the earlier study had lower statistical powerthan the present study

Exploratory Questions

Variance Measure

Table 7 shows the results for the exploratory measure of EDA variance lt 50represents reduced variance gt 50 represents increased variance We found nodifference between variance scores for cooperate versus block conditions(t = 150 df = 49) When these are combined the resulting figure does not de-viate from chance expectation (t = 440 df = 49) Thus as in Study 1 it seems

TABLE 6Summary Data Mean PSI Scores

(N = 50 MCE = 50 range in brackets)

Co-operate PSI Block PSI Combined PSI

5008 5019 5016(4821- 5221) (4807- 5557) (4900- 5296)

that the exploratory variance measure does not reveal any additional indica-tion of the operation of a remote mental influence in this study However thewider range of variance scoring continues to suggest that this measure mayhave greater potential sensitivity than the summary PSI score

Influenceesrsquo Blocking and Co-operating Strategies

We were interested in learning more about what particular strategies partici-pants chose to adopt when blocking and co-operating As in Study 1 a post-session questionnaire was used to gather this information Table 8 shows thepercentage of participants who adopted each kind of strategy

The table shows a very similar pattern to that found in Study 1 The mainpoint of interest is that almost no participants took a passive approach toblocking (ie adopting an openrelaxing strategy) A much higher proportionpreferred to adopt this approach when co-operating The preferred mode forboth conditions was to use some form of visual imagery such as imagining ashield or a screen when blocking and imagining an open channel such as ariver when co-operating

The majority of participants reported no concerns over the possibility ofhaving their physiology remotely influenced by another person as found instudy one

Sex Effects

As in study one the sexes were not equally represented among participantsWhereas Study 1 had an excess of male participants in Study 2 twice as manyfemales (N = 33) as males (N = 17) participated It was therefore important tolook for sex differences since any such differences could confound the inter-pretation of overall trends We found no difference between sexes when co-op-erating (female mean PSI score = 5008 male mean PSI score = 5006 t = 107 df = 48) There was a nonsignificant trend for females to have higher

Blocking vs Co-operating 529

TABLE 7Summary Data Mean Variance Scores(N = 50 MCE = 50 range in brackets)

Co-operate Variance Block Variance Combined Variance

5076 5054 5034(3164- 7279) (3572- 6813) (3749- 6671)

TABLE 8Influenceesrsquo Mental Strategies ()

Visual Imagery Cognitive VisCog OpenRelax

Blocking 64 26 6 2Co-operating 60 8 12 20

530 C Watt et al

PSI scores than males when blocking (mean PSI scores 5030 and 4998 re-spectively t = 992 df = 48) Combining blocking and co-operating conditionsthere was no overall difference between the sexes (mean PSI scores 5020 forfemales 5008 for males t = 665 df = 48)

Experimenter Effects

Study 2 found a statistically significant overall remote influence effect(combined PSI = 50155 t = 1806) Table 9 shows the breakdown of remoteinfluence scores for the two experimenters CW and JR Looking at the overalldata for the two experimenters we see that CWrsquos sessions were associatedwith most of the remote influence effect However due to the smaller N CWrsquossessions are not independently significant (t = 1632) and there is not a signifi-cant difference between CW and JRrsquos overall PSI scores (t = 678) Post hocwe calculated effect size r according to the formula r = [t2(t2+df)]12 for the twoexperimenters For JR (t = 817 df = 22) we found a small effect sizer = +17 For CW (t = 1632 df = 26) the effect size was moderate r = +30This may be some post hoc indication of an experimenter effect in Study 2

InfluenceeInfluencer Sex Pairings

As can be seen from Table 10 Study 2 found exactly the opposite trend tothat found in Study 1 On a post hoc test same sex pairings scored significant-ly higher on the remote influence PSI measure than mixed sex pairings(t = - 2034 df = 48 p = 048 2 - t) This contradicts Schlitz and LaBergersquos(1997) finding of greater autonomic detection of remote staring for mixed sexpairs

TABLE 9Remote Influence Results by Experimenter

Experimenter N sessions Coop PSI Block PSI Combined PSI

CW 27 5011 5026 5021JR 23 5003 5010 5009

TABLE 10Mean PSI Scores for Different InfluenceeInfluencer Sex Pairings

Pairing N Coop PSI Block PSI Combined PSI

MF 11 4977 5009 4997FM 17 4983 5021 5003FF 16 5035 5038 5038MM 6 5060 4978 5027

General Summary and Conclusions

We have described the findings of two studies exploring the limits of remotemental influence effects We wished to see whether the mental strategiesadopted by the influencee would affect remote influence scores In particularcould an influencee block or resist a remote influence attempt We recorded in-fluenceesrsquo electrodermal activity while the influencer attempted to calm to ac-tivate or to have no influence over the influenceesrsquo physiology In each studythere were two conditions reflecting the influenceesrsquo designated mental strate-gies in response to the remote influence attempt blocking and co-operating

There were only two formal hypotheses 1) that remote influence scoreswould be significantly higher during the co-operating condition than duringthe blocking condition and 2) that overall there would be significant evidenceof remote mental influence on physiology in the designated direction (iecalming when calming was the influencerrsquos intention activating when activat-ing was the influencerrsquos intention) Contrary to prediction Study 1 found thatremote influence effects were higher in the block condition than in the co-op-erate condition (but not to a significant extent) Also overall remote influenceperformance was at chance level This study therefore found no support for ei-ther of the formal hypotheses Study 2 consisted of a replication of Study 1with a few minor adjustments plus additional measures that were not reportedin this paper Again contrary to prediction remote influence scores were non-significantly higher in the block condition than in the co-operate conditionThis replicates the trend found in Study 1 failing to support hypothesis oneThese results suggest that the mental strategy adopted by the influencee in theface of a remote influence attempt has no effect on remote influence perfor-mance Referring back to the questions posed in the introduction then wehave not found any evidence to suggest that it is possible to resist a remote in-fluence effect and there is as yet no indication of the possible limiting condi-tions for remote mental influence

Unlike Study 1 Study 2 found statistically significant evidence of an overallremote influence effect in the prespecified direction (ie calming of EDAwhen that was the influencerrsquos intention activation of EDA when that was theinfluencerrsquos intention) In order to facilitate comparison with a recent meta-an-alytic review of remote influence studies we calculated the effect size (r) ac-cording to the formula r = [t2(t2+df)]12 For their review of 19 EDA influencestudies Schlitz and Braud (1997) found an average effect size of r = +25 OurStudy 1 with overall t = - 031 df = 31 gives a negligible effect size of r = - 006 Study 2 with overall t = 1806 df = 49 gives r = +25 Study 2 there-fore replicates the mean effect size found in EDA studies conducted by otherresearchers in other parapsychology laboratories Post hoc we can combinethe data of the two studies for greater statistical power Overall t = 1176 df = 81 giving a small effect size of r = +13

We can find no clear reason why Study 2 found an EDA influence effectwhile Study 1 did not A number of factors are worth considering 1) The

Blocking vs Co-operating 531

532 C Watt et al

significant result could be due to chance or undetected artifact 2) Study 2 im-plemented what we considered to be minor methodological improvements thatmight have removed some artifactual fluctuations from the EDA data that weobserved in study one 3) Each experimenter in Study 2 conducted three testingsessions with their participant By the third session in which the remote influ-ence measure was taken experimenters were on quite friendly terms with theirparticipants perhaps more so than in Study 1 where each participant came infor only one testing session 4) The participants in Study 2 were more represen-tative of the general population we felt than those in Study 1 Study 1 in-volved more undergraduate psychology students and associates of the experi-menters 5) The experimenters in study one were less experienced with theremote influence protocol than when they conducted Study 2

In addition to the formal hypotheses we reported a number of exploratorymeasures Study 1 found that influenceesrsquo self-reported vividness of visual im-agery appeared to be unrelated to the remote influence performance In bothstudies influenceesrsquo qualitative descriptions of the actual mental strategiesthey employed when blocking and co-operating revealed a tendency for differ-ent strategies to be preferred for the different conditions The use of visual im-agery was popular for both blocking (eg visualizing being surrounded by aprotective cocoon) and co-operating (eg visualizing a river flowing betweeninfluencer and influencee) However a considerable proportion of participantsreported adopting an open or relaxed frame of mind when co-operating whilstalmost no participants chose this strategy when blocking This was despite thefact that prior to each session we encouraged participants to adopt comparablestrategies for blocking and co-operating

It was interesting and perhaps surprising to note that in both studies themajority of participants reported no feelings of concern or discomfort at thepossibility of having their physiology remotely influenced Those who wereconcerned tended to be those who knew more about parapsychology and thepossible harmful applications of remote influence A small proportion of par-ticipants was unconcerned because they did not believe such an influence waspossible in the first place

Both studies employed an exploratory variance measure of EDA influenceThis measure showed a greater range of scoring overall thus perhaps greaterpotential sensitivity to a remote influence effect However the variance mea-sure showed no indication of a remote influence effect To date researchershave employed a variety of indicators of remote influence as the literaturegrows progress may be made on identifying which is the most appropriate in-dicator

Neither study found significant sex differences in remote influence effectsthough in both studies female influencees tended to have higher remote influ-ence scores than male influencees

In terms of their overall remote influence effects there appeared to be nodifference between experimenterinfluencers in Study 1 (r = +06 - 05 and

- 06) In Study 2 effect size measures indicated that the experimenters dif-fered in remote influence effect sizes (r = +17 and +30) Due to the smaller Nthis difference was not found to be statistically significant Post hoc combin-ing Studies 1 and 2 for experimenters JW and CW (ZMcD was not an experi-menter in Study 2) the effect sizes are r = +08 and r = +20 respectively

We conducted a post hoc analysis of the effects of influenceeinfluencer sexpairings on remote influence scores Our two studies came up with contradic-tory findings each statistically significant Study one found that mixed sexpairings had significantly higher remote influence scores than same sex pair-ings This replicated the post hoc finding of Schlitz and LaBerge (1997) in con-nection with the autonomic detection of remote observation Study 2 foundthat same sex pairings had significantly higher remote influence scores thanmixed sex pairings Our contradictory findings do not help to clarify the pic-ture on the importance or otherwise of sex pairings It is quite possible that sexis not the most salient aspect of the dynamic between pairs Age personalityand familiarity are likely to be important variables moderating the relationshipbetween influencer and influencee

There are many more variables that we could look at but as the number ofanalyses multiplies so does the opportunity for spurious significant resultsFor this reason we have named only two formal hypotheses for each experi-ment All other analyses are regarded as exploratory and any trends that wehave highlighted above would require replication

In conclusion these two studies have found no evidence of a difference inremote influence effects when influencees are asked to adopt cooperative men-tal strategies compared to when they are asked to adopt blocking mental strate-gies In fact contrary to expectation both studies found slightly higher remoteinfluence effects for the blocking condition compared to the co-operating con-dition On this basis we have found no indication of limiting conditions for re-mote influence effects Therefore while anecdotal advice exists on how to de-fend oneself against a remote physiological influence attempt in the realworld laboratory evidence to date has provided no support for this advice

Given the importance of this question both in practical terms and in theoret-ical terms we argue that this line of research deserves continued investigationIf a way could be found to conduct research with a more realistic scenario mdash where the remote influence was not just neutrally and abstractlydefined as unwelcome mdash then our research might have more meaningful thingsto say about the parameters of remote influence It is interesting and perhapssurprising to note that when Watt has lectured on this topic to informed lay au-diences the feedback has often been that researchers must use more potential-ly harmful or disruptive remote influence (of course with the informed consentof the participants) While there are clear ethical inhibitions when consideringpossible physiological influence perhaps research using higher level influencewould be more acceptable For example one could attempt to disrupt or en-hance the concentration of chess players who were aware of the influence

Blocking vs Co-operating 533

534 C Watt et al

attempt and who were concurrently blocking or co-operating with the influ-ence attempt Research already exists to suggest that remote influence does notjust operate at physiological levels Attention focusing for example can be fa-cilitated by a remote helper (Brady amp Morris 1997 Braud et al 1995) Futureresearch in this area must therefore expand beyond physiological influences

Acknowledgements

The authors gratefully acknowledge the financial support they receivedfrom the Institut fuumlr Grenzgebiete der Psychologie und PsychohygieneFreiburg Germany while conducting these studies We would also like tothank Dr Paul Stevens for generously helping to solve software difficultieswhen preparing for these studies and the anonymous reviewer for helpfulcomments Address reprint requests to Dr C Watt Department of PsychologyUniversity of Edinburgh 7 George Square Edinburgh EH8 9JZ Scotland UK

ReferencesBrady C amp Morris R L (1997) Attention focusing facilitated through remote mental interac-

tion A replication and exploration of parameters In R Wiseman (Ed) Proceedings of Pre-sented Papers The Parapsychological Association 40th Annual Convention held in conjunc-tion with The Society for Psychical Research pp 73ndash91

Braud W G (1985) Blockingshielding psychic functioning through psychological and psychictechniques A report of three preliminary studies In RA White and J Solfvin (Eds) Re-search in Parapsychology 1984 Metuchen NJ Scarecrow Press pp 42ndash44

Braud W G amp Schlitz M J (1991) Consciousness interactions with remote biological systemsAnomalous intentionality effects Subtle Energies 2 1

Braud W G Schlitz M J Collins J and Klitch H (1985) Further studies of the bio-PK effectFeedback blocking specificitygenerality In RA White amp J Solfvin (Eds) Research inParapsychology 1984 Metuchen NJ Scarecrow Press pp 45ndash48

Braud W G Shafer D McNeill K and Guerra V (1995) Attention focusing facilitatedthrough remote mental interaction Journal of the American Society for Psychical Research89 103

Cohen J (1997) Statistical Power Analysis for the Behavioral Sciences New York AcademicPress

Dalton K (1994) Remotely influenced ESP performance in a computer task A preliminarystudy Proceedings of the 37th Annual Convention of the Parapsychological Association pp95ndash103

Dossey L (1994) Healing and the mind Is there a dark side Journal of Scientific Exploration 873

Halifax-Grof J (1974) Hex death In A Angoff amp D Barth (Eds) Parapsychology and Anthro-pology Proceedings of an International Conference Held in London England August 29ndash311973 NY Parapsychology Foundation

Hope M (1983) Practical Techniques of Psychic Self-Defence Wellingborough Northampton-shire Aquarian Press

Marks DF (1973) Visual imagery differences in the recall of pictures British Journal of Psy-chology 64 17

Nash CB (1982) Psychokinetic control of bacterial growth Journal of the Society for PsychicalResearch 51 217

Radin DI Taylor RK and Braud WG (1995) Remote mental influence of human electroder-mal activity A pilot replication European Journal of Parapsychology 11 19

Schlitz M amp Braud W G (1997) Distant intentionality and healing Assessing the evidence Al-ternative Therapies 3 62

Schlitz M J amp LaBerge S (1997) Covert observation increases skin conductance in subjectsunaware of when they are being observed A replication Journal of Parapsychology 61 185

Schmeidler G R (1958) Agent-percipient relationships Journal of the American Society forPsychical Research 52 47

Schmeidler G R (1961) Evidence for two kinds of telepathy International Journal of Parapsy-chology 3 5

Tart C T (1979) A survey of expert opinion on potentially negative uses of psi United Statesgovernment interest in psi and the level of research funding of the field In WG Roll (Ed) Re-search in Parapsychology 1978 Metuchen NJ Scarecrow Press pp 54

Tart C T amp Labore C M (1986) Attitudes towards strongly functioning psi A preliminary sur-vey Journal of the America Society for Psychical Research 80 163

Tart C T (1986) Psychicsrsquo fears of psychic powers Journal of the American Society for Psychi-cal Research 80 278

Watt C A McDermott Z amp Ravenscroft J (1997) Exploring the limits of remote mental inter-action A comparison of blocking versus co-operating strategies In R Wiseman (Ed) Pro-ceedings of the 40th Annual Convention of the Parapsychological Association held in conjunc-tion with the Society for Psychical Research Brighton England pp 464ndash476

Blocking vs Co-operating 535

analysis Their function was primarily to separate the calm and activateepochs from one another and to separate any lingering influence that may havebeen left at the end of the influencerrsquos calm or activate epochs In the absenceof any remote influence effect PSI equals 50 A PSI of greater than 50 rep-resents a remote influence effect in the predicted direction (ie calming whenthat was the influencerrsquos intention activating when that was the influencerrsquosintention) A PSI of less than 50 represents a remote influence effect in thedirection opposite to that predicted For the present study we calculated PSIseparately for each participantrsquos blocking and co-operating condition plus wecalculated an overall PSI score for each participant

Participants

The influencees were volunteer participants drawn from the authorsrsquo con-tacts in the Department of Psychology at Edinburgh University and includingundergraduate psychology students recruited by e-mail and by word of mouthThere were three experimentersinfluencers overall (Watt Ravenscroft andMcDermott) but in each testing session there were only two individuals mdash thevolunteer influencee and the influencer Each experimenter also took part asinfluencer in one session

Written Materials

Each volunteer was given a one page information sheet Once they had readthis they were given a 34-item participant information form (PIF) and a 16-item Vividness of Visual Imagery Questionnaire (Marks 1973) After dis-cussing the procedure with the influencer each participant was asked to sign aconsent form that assured confidentiality and noted their right to withdrawfrom the study at any time Following completion of the DMILS procedure in-fluencee and influencer both completed one-page qualitative post-sessionquestionnaires in which they were invited to describe the kinds of strategiesthat they used during the testing session and to give their subjective impres-sions of the session A letter of thanks and a brief overview of the results weregiven to participants at the end of the study

Technical Apparatus

Two rooms were used for the testing sessions The influenceersquos room wassound-attenuated with double-doors and was located approximately 20 metersaway from the influencerrsquos room down a short flight of stairs The DMILS ap-paratus (an I-410 multichannel physiological monitoring unit from J amp J Sys-tems with source code from Physiodata) was set up between these two roomsand was run on two time-synchronized Pentium computers In the influenceersquosroom a monitor displayed written instructions and a mildly engaging visualdisplay (a screen saver shown during the actual data collection epochs) to theinfluencee This room also contained speakers that gave auditory instructions

Blocking vs Co-operating 519

520 C Watt et al

(ldquohey block nowrdquo ldquohey co-operate nowrdquo and ldquosession over thank yourdquo) tothe influencee just prior to each of the two 10-minute runs duplicating thewritten instructions that appeared on the computer monitor The physiologysensors that were attached to the influenceersquos fingers were located in the influ-enceersquos room The signals from the sensors were relayed along a cable andprocessed by a computer located in the influencerrsquos room Therefore the com-puter in the influenceersquos room was merely used to provide instructions and avisual display to the influencee The computer in the influencerrsquos room con-trolled the DMILS software and processes and stored the psychophysiologydata

Procedure

The influencer and participant initially sat down in the reception room in theparapsychology suite for a further description of procedure and returned theircompleted PIFimagery questionnaire to the influenceer who went over it withthe participant Any questions that were raised by the participant prior to test-ing were answered here (This question-and-answer session usually lasted fif-teen minutes) The participant then gave their informed consent to the studyThe influencer also suggested various mental approaches the participant mightadopt for the blocking and co-operating conditions (some examples weregiven in the information sheet) These suggestions included both visual im-agery (eg visualizing a shielding cocoon in the blocking condition) and cog-nitive strategies (eg adopting a stubborn and uncooperative frame of mind inthe blocking condition) Participants were informed that different strategiesmight suit different people and they were to do whatever felt natural for themThe influencer then took the participant to the testing room and demonstratedthe auditory instructions so that the participant would not be startled by thesewhen they occurred during the session itself The participant was also taken tosee where the influencer was sitting and the influencer described what theywould be doing when they were attempting to activate or calm the participant

Once the preliminaries were finished the influencee was seated in a partiallyreclined chair in front of the computer monitor and the physiology sensorswere attached to their fingers by the influencer The room was softly illuminat-ed by a desk lamp with an orangered filter At this stage the influencer checkedthat the physiology signals were being transmitted to the computer in theirroom Returning to the influencee the influencer set a start time on the influ-enceersquos computer This was on average two to three minutes later The influ-encer then returned to their room closing all doors and waited for the start ofthe session

Physiology Data Collection As the session began the influencee was in-structed visually and auditorially either to ldquohey block nowrdquo or ldquohey co-oper-ate nowrdquo The influenceersquos monitor then displayed a colorful and dynamicscreen-saver that was intended to be engaging without being particularlyarousing or relaxing At the same time the influencer was following the influ-

ence schedule as instructed by the computer observing the real-time feedbackdisplay of the participantrsquos EDA when desired At the end of the first run whichlasted approximately ten minutes the influencer saw a 150 second count downon the computer before the second run began At the same time the influenceesaw a message about the strategy to adopt for the forthcoming second run (ieto block or to co-operate whichever strategy was not used for the first run) Atthe beginning of the second run the influencee was also instructed auditoriallyeither to ldquohey block nowrdquo or ldquohey co-operate nowrdquo The dynamic visual dis-play resumed while the influencer followed a random influence schedule forapproximately ten minutes The entire session took about 25 minutes to com-plete At the end of the session the influencee was auditorially and visually in-formed ldquosession over thank yourdquo and waited until the influencer returned tothe influenceersquos room

Post-Testing Procedures The influenceersquos physiology sensors were re-moved by the influencer and both returned to the reception room to completethe post-session questionnaires Then the influencee was taken to the influ-encerrsquos room to see feedback on the results of their testing session There wasan option for some final discussion about the session after which the volun-teer was thanked for their participation and left Once all the data was collect-ed feedback on the results of the study as a whole was sent to participants

Planned Hypotheses and Analyses

Our main aim was to examine whether there was any evidence for blockingof a remote mental influence on physiology [EDA is the principal physiologymeasure here with per-session PSI scores representing the measure of theDMILS effect on EDA] H1 therefore predicted significantly greater PSIscores in the co-operating condition than in the blocking condition (using a re-lated t-test one-tailed) Our second question was whether there was any over-all evidence of remote mental influence on EDA This analysis would be con-ducted on the data of blocking and co-operating runs combined if nosignificant difference was found between these two conditions (as tested byH1) If H1 was supported H2 would be tested on the data of co-operating runsonly predicting PSI scores significantly greater than chance expectation(using a single-mean t-test of the PSI score vs MCE one-tailed)

Planned alpha level was 05 There was no correction for multiple analysisThis was dealt with by restricting the number of formal hypotheses (with one-tailed significance tests) to two in each study All other analyses were regardedas exploratory and two-tailed significance tests were used

Exploratory Measures

Variance

DMILS researchers do not yet appear to have identified what might be themost appropriate indicator of a remote influence effect Therefore in addition

Blocking vs Co-operating 521

522 C Watt et al

to the PSI score we looked at the amount of variance that was shown in block-ing and co-operating scores The variance measure was calculated as followsMean values for activate periods and calm periods were calculated separatelyfor block and co-operate conditions Mean values were taken from each indi-vidual score in that condition and squared and summed S [mean value for con-dition - individual score]2 This gave a summation of variability around themean for each activate calm and rest period for block and cooperate condi-tions It was used in conjunction with the mean level of EDA activity givingan indication of the motion of the wave (ie the variability of the EDA activi-ty) rather than the overall level

VVIQ

Anecdotally the use of appropriate imagery (eg visualize being surroundedby a shielding white halo of white light) is suggested to be an effective blockingtechnique (eg Hope 1983) We therefore wished to see if participants who re-ported experiencing vivid imagery with ease are perhaps more effective block-ers than those who do not report clear mental imagery Descriptive data wasgathered to explore any trends in PSI scores for blockingco-operating forthose participants with VVIQ scores ranging from low (ie reporting vividmental imagery) to high (ie reporting little or no mental imagery)

Qualitative Reports

We also wanted to report data illustrating what kind of different strategiesparticipants used when they were blocking and co-operating These qualitativereports were gathered in the form of the participantsrsquo written responses toopen-ended questions on a post-session questionnaire Watt and Ravenscroftindependently coded these written responses into categories and cross-checked to ensure coding agreement

Sex and Experimenter Effects

We also examined the data for sex effects and experimenter effects becausethese variables could confound the interpretation of any overall effects Final-ly a post hoc analysis was conducted on influenceeinfluencer sex pairings Ina study on the unconscious detection of remote staring Schlitz and LaBerge(1997) found the greatest effect for opposite-sex observer-subject pairs

Study One Results and Discussion

Participants

Thirty-two individuals took part as influencees (12 females 20 males) theirages ranged from 19 to 63 years (mean age 285 years)

Incomplete Data

One participant declined to give their age on the PIF form so the descriptivestatistics on age do not include this participant (so N = 31) Also due to a tech-nical hitch the exploratory variance scores were lost for one participant so N = 31 for this measure There were no drop-outs from this study

Remote Influence Measure

Table 1 presents a summary of the remote influence data in the form ofmean Percentage Score Index for co-operating and blocking conditions Therewas no significant difference between PSI for co-operating and blocking con-ditions (t = - 202 df = 31) Therefore Hypothesis 1 (predicting significantlygreater PSI scores for co-operating condition compared to blocking condition)was not supported by these results As planned the PSI scores for co-operatingand blocking runs were combined to examine Hypothesis 2 (of an overall re-mote mental influence effect in line with the calming and activating intentionsof the influencee) Mean PSI scores were exactly at chance expectation (t = - 031 df = 31) therefore there was no support for Hypothesis 2

Exploratory Questions

Variance Measure

Table 2 presents summary data for the exploratory measure of the amount ofvariance shown in each condition (lt 50 represents reduced variance gt 50 rep-resents increased variance) There was no significant difference between vari-ance scores for co-operating and blocking conditions (t = 835 df = 30) andwhen the conditions were combined variance did not depart from chance ex-pectation (t = - 244 df = 30) It is interesting to note however that the vari-ance measure gives a much wider range of scores than the PSI measure

Blocking vs Co-operating 523

TABLE 1Summary Data Mean PSI Scores

(N = 32 MCE = 50 range in brackets)

Co-operate PSI Block PSI Combined PSI

4996 5003 5000(4692- 5215) (4653- 5404) (4853- 5309)

TABLE 2Summary Data Mean Variance Scores(N = 31 MCE = 50 range in brackets)

Co-operate Variance Block Variance Combined Variance

5089 4864 4977(2648- 7012) (2492- 6484) (4116- 6297)

524 C Watt et al

suggesting that the variance measure may potentially be a more sensitive indi-cator of remote influence effect than PSI

Vividness of Mental Imagery

Low VVIQ scores represent participants reporting highly vivid imagery andhigh VVIQ scores represent those reporting little or no imagery The meanscore was 4275 (range from 21 to 80) Using a Pearson correlation we foundthat there was no evidence to suggest that those participants with low VVIQscores (high imagery) were better blockers than those with high scores (VVIQvs PSI blocking r = - 113) We also found that there was no evidence of a cor-relation between vividness of imagery and overall remote influence effectsusing the combined Percentage Score Index measure (VVIQ vs PSI overall r =- 077)

Influenceesrsquo Blocking and Co-operating Strategies

Influencees completed a questionnaire (after the session was over) askingthem to describe what strategies they used for blocking and co-operating Cat-egorizing these responses we found a difference in the kinds of strategies usedwhen blocking compared to co-operating as shown in Table 3 In the blockingcondition the majority of participants (469) used visual imagery strategiesFewer (281) participants used a combination of imagery and cognitivestrategies (such as repeating words related to blocking) while 25 reportedusing purely cognitive strategies for blocking (such as adopting a resistant anduncooperative attitude or thinking of irrelevancies) In the co-operating con-dition fewer participants used visual imagery alone (333) The majority ofparticipants (424) reported that their strategies for co-operating includedadopting a feeling of passiveness openness or relaxation whereas no partici-pants adopted this kind of strategy for the blocking condition 121 of partic-ipants used cognitive strategies alone (such as adopting a co-operative atti-tude) and the same proportion used a combination of visual imagery andcognitive strategies

Therefore there was a tendency for participants to use visual imagery forblocking and relaxationpassivity or visual imagery for co-operating This wasdespite the fact that we asked participants to attempt to adopt similar levels ofactivation for both co-operating and blocking strategies There was also a ten-dency not surprisingly for those participants with low VVIQ scores (ie vivid

TABLE 3Influenceesrsquo Mental Strategies ()

Vis Im Cognitive VisCog OpenRelax

Blocking 469 250 281 00Co-operating 333 121 121 424

imagers) to use visual imagery strategies Those who had high VVIQ scoresalso used visual imagery but more often used this in combination with othertactics

Other Qualitative Reports from Participants According to their post-ses-sion questionnaire responses participants generally felt comfortable about theattempt to be influenced (N = 24) Three participants said they felt apprehen-sive or uncomfortable with the idea of being influenced Interestingly of thesethree two were actively involved with parapsychology and were aware of theharmful psi implications discussed at the beginning of this report The thirdwas a psychology student who had attended a lecture given by Watt in whichthe possibilities of harmful psi were discussed Another group of three partici-pants said they would be uncomfortable with the idea of being influenced ifthey believed such influence was possible

Seven participants said they found it was difficult to sustain concentrationfor the 10 minutes that comprised each condition Five participants found thescreensaver distracting or disturbing Two participants specifically mentionedthat they found blocking a more taxing task than co-operating

Sex Effects

There were more males (N = 20) than females (N = 12) in this study howev-er there was no significant difference in PSI scores for remote physiology in-fluence between males and females (male PSI vs female PSI t = 1085 male PSI = 4987 female PSI = 5021)

Experimenter Effects

We studied whether there was any difference in PSI scores obtained by thethree experimenters The results are shown in Table 4 There seems to be nosign of experimenter effects in study one

InfluenceeInfluencer Sex Pairings

Table 5 gives the mean PSI scores for different influenceeinfluencer pair-ings Taking the combined PSI scores it can be seen that the highest remote in-fluence effect is for FM pairs followed by MF pairs followed by FF pairsand the lowest scoring is for MM pairs Post hoc it was found that the mixedsex pairings showed significantly higher remote influence scores than the same

Blocking vs Co-operating 525

TABLE 4Remote Influence Results by Experimenter

Experimenter N sessions Coop PSI Block PSI Combined PSI

CW 12 4983 5011 4997ZMcD 10 5015 4998 5006JR 10 4994 4999 4996

526 C Watt et al

sex pairings (t = - 2016 df = 30 p = 053 2 - t) This replicates the effect foundby Schlitz and LaBerge (1997)

Study One Conclusions

In Study 1 we developed a methodology to examine intentional resistance toa remote mental influence attempt (Watt et al 1997) We contrasted influ-enceesrsquo physiological activity when they were mentally co-operating with aremote attempt to influence their physiology with physiological activity wheninfluencees were mentally blocking the influence attempt We found no evi-dence for any remote influence effect overall and no difference betweenblocking and co-operating conditions Our qualitative measures yielded someinteresting information about participantsrsquo mental strategies and reactions dur-ing the remote influence session We found no difference between experi-menters in remote influence scores but we did find that mixed sex influ-enceeinfluencer pairings scored more highly than same sex pairings

Study Two Method

Study 2 was planned to explore in more depth the nature of the relationshipbetween aspects of personality and cognition and psi performance Partici-pants took part in three testing sessions approximately one week apart Thefirst consisted of various psychological measures the second consisted of aforced choice ESP task and the third consisted of a replication of the blocking-co-operating study (Study 1) described above The psychological measureswere not scored until after completion of the remote influence session Partici-pants received feedback about their forced choice ESP performance at the con-clusion of the second session For ease of interpretation the present paper isrestricted to a description of the third testing session which was essentially areplication of study one with fewer influencers and greater statistical power

We introduced some slight changes to the methodology of Study 2 based onour experiences with Study 1 At the start of each of the two runs in Study 1the participant was given auditory instructions either to ldquoblock nowrdquo or ldquoco-operate nowrdquo We observed that the auditory instructions appeared to affect theparticipantrsquos EDA as if they were startled As planned the first recordingepoch in each run was not used in analysis to give the participantrsquos EDA time

TABLE 5Mean PSI Scores for Different InfluenceeInfluencer Sex Pairings

Pairing N Coop PSI Block PSI Combined PSI

M-F 6 4932 5069 5001F-M 6 5043 5034 5048F-F 6 5022 4952 4993M-M 14 4988 4975 4981

to settle down However in order to remove any remaining potential artifactu-al effects on EDA in Study 2 only visual instructions were given

Another difference in methodology concerned the screensaver viewed bythe participants during the experimental session In Study 1 some participantsreported that the series of patterns produced by the computer were being dis-played at too fast a rate such that some participants found it distracting or dis-turbing Again we felt that this may have an artifactual effect on EDA so wedecided for study two to reduce the speed of the screensaver

Thirdly in Study 2 we did not give participants detailed information abouttheir remote influence performance until the feedback letters that were distrib-uted at the conclusion of the entire study The reason for this change was mere-ly to streamline the procedure on what was already a time-consuming experi-ment Finally in another streamlining change we did not take VVIQ measuresin Study 2

Planned Hypotheses and Analyses

These were the same as with Study 1 That is hypothesis 1 predicted higherPSI scores for the co-operating condition than for the blocking condition(measured with a related t-test one-tailed) If no significant difference wasfound between blocking PSI and co-operating PSI scores the two would becombined to look at overall remote influence effect (hypothesis 2 single-meant-test one-tailed) If a significant difference was found with hypothesis 1 theanalysis for hypothesis 2 would be conducted with the data for co-operatingsessions only

Exploratory Questions

As with study one we took an exploratory variance measure of remote EDAinfluence We also gathered qualitative descriptions of the mental strategiesparticipants adopted whilst blocking and co-operating We also looked at sexeffects experimenter effects and post hoc influenceeinfluencer sex pairings

Participants

Study 2 had two experimenterinfluencers (Watt amp Ravenscroft) After aparticipant was recruited (usually by Ravenscroft) they were assigned to oneof the two experimenters for all three testing sessions This assignment wasdone unsystematically according to the preferences and availability of the ex-perimenters As planned 50 volunteer participants were drawn from the gener-al participant pool at the Koestler Parapsychology Unit and by word ofmouth All participants completed the Participant Information Form and aConsent form prior to the study The participants consisted of 33 females and17 males Average age was 41 years range from 22 years to 76 years of age Inorder to increase the studyrsquos relevance to the general population undergradu-ate psychology students were not recruited as participants The main criterion

Blocking vs Co-operating 527

528 C Watt et al

for participation was that the participants were interested in and open-mindedabout parapsychology Again there were no drop-outs from this study Allparticipants received a detailed personal feedback letter at the conclusion ofthe study

Study Two Results and Discussion

Remote Influence Measure

PSI scores are summarized in Table 6 which shows that the mean PSI scorewhen co-operating was nonsignificantly lower than mean PSI score whenblocking (t = - 595 df = 49) Hypothesis 1 was therefore not supported Asplanned in the absence of any significant difference between blocking and co-operating PSI scores the two conditions were combined to test hypothesis 2which predicted overall significant remote mental influence performance Thecombined PSI score was significantly greater than mean chance expectation of50 (t = 1806 df = 49 p = 038 1 - t) Therefore hypothesis 2 was supportedgiving evidence of an overall remote influence effect upon psychophysiologyin the prespecified direction

The failure to support hypothesis 1 confirms what we found in Study 1 thatthere was no significant difference between blocking and co-operating condi-tions on remote influence scores Indeed both studies found a slight effect inthe opposite direction to that predicted that is slightly higher PSI scores whenblocking was the influenceersquos intention than when co-operating was their in-tention It was encouraging to find significant support for a remote influenceeffect in the predicted direction for blocking and co-operating conditions com-bined No such trend was found in the previous study though with a smallernumber of participants (N = 32) the earlier study had lower statistical powerthan the present study

Exploratory Questions

Variance Measure

Table 7 shows the results for the exploratory measure of EDA variance lt 50represents reduced variance gt 50 represents increased variance We found nodifference between variance scores for cooperate versus block conditions(t = 150 df = 49) When these are combined the resulting figure does not de-viate from chance expectation (t = 440 df = 49) Thus as in Study 1 it seems

TABLE 6Summary Data Mean PSI Scores

(N = 50 MCE = 50 range in brackets)

Co-operate PSI Block PSI Combined PSI

5008 5019 5016(4821- 5221) (4807- 5557) (4900- 5296)

that the exploratory variance measure does not reveal any additional indica-tion of the operation of a remote mental influence in this study However thewider range of variance scoring continues to suggest that this measure mayhave greater potential sensitivity than the summary PSI score

Influenceesrsquo Blocking and Co-operating Strategies

We were interested in learning more about what particular strategies partici-pants chose to adopt when blocking and co-operating As in Study 1 a post-session questionnaire was used to gather this information Table 8 shows thepercentage of participants who adopted each kind of strategy

The table shows a very similar pattern to that found in Study 1 The mainpoint of interest is that almost no participants took a passive approach toblocking (ie adopting an openrelaxing strategy) A much higher proportionpreferred to adopt this approach when co-operating The preferred mode forboth conditions was to use some form of visual imagery such as imagining ashield or a screen when blocking and imagining an open channel such as ariver when co-operating

The majority of participants reported no concerns over the possibility ofhaving their physiology remotely influenced by another person as found instudy one

Sex Effects

As in study one the sexes were not equally represented among participantsWhereas Study 1 had an excess of male participants in Study 2 twice as manyfemales (N = 33) as males (N = 17) participated It was therefore important tolook for sex differences since any such differences could confound the inter-pretation of overall trends We found no difference between sexes when co-op-erating (female mean PSI score = 5008 male mean PSI score = 5006 t = 107 df = 48) There was a nonsignificant trend for females to have higher

Blocking vs Co-operating 529

TABLE 7Summary Data Mean Variance Scores(N = 50 MCE = 50 range in brackets)

Co-operate Variance Block Variance Combined Variance

5076 5054 5034(3164- 7279) (3572- 6813) (3749- 6671)

TABLE 8Influenceesrsquo Mental Strategies ()

Visual Imagery Cognitive VisCog OpenRelax

Blocking 64 26 6 2Co-operating 60 8 12 20

530 C Watt et al

PSI scores than males when blocking (mean PSI scores 5030 and 4998 re-spectively t = 992 df = 48) Combining blocking and co-operating conditionsthere was no overall difference between the sexes (mean PSI scores 5020 forfemales 5008 for males t = 665 df = 48)

Experimenter Effects

Study 2 found a statistically significant overall remote influence effect(combined PSI = 50155 t = 1806) Table 9 shows the breakdown of remoteinfluence scores for the two experimenters CW and JR Looking at the overalldata for the two experimenters we see that CWrsquos sessions were associatedwith most of the remote influence effect However due to the smaller N CWrsquossessions are not independently significant (t = 1632) and there is not a signifi-cant difference between CW and JRrsquos overall PSI scores (t = 678) Post hocwe calculated effect size r according to the formula r = [t2(t2+df)]12 for the twoexperimenters For JR (t = 817 df = 22) we found a small effect sizer = +17 For CW (t = 1632 df = 26) the effect size was moderate r = +30This may be some post hoc indication of an experimenter effect in Study 2

InfluenceeInfluencer Sex Pairings

As can be seen from Table 10 Study 2 found exactly the opposite trend tothat found in Study 1 On a post hoc test same sex pairings scored significant-ly higher on the remote influence PSI measure than mixed sex pairings(t = - 2034 df = 48 p = 048 2 - t) This contradicts Schlitz and LaBergersquos(1997) finding of greater autonomic detection of remote staring for mixed sexpairs

TABLE 9Remote Influence Results by Experimenter

Experimenter N sessions Coop PSI Block PSI Combined PSI

CW 27 5011 5026 5021JR 23 5003 5010 5009

TABLE 10Mean PSI Scores for Different InfluenceeInfluencer Sex Pairings

Pairing N Coop PSI Block PSI Combined PSI

MF 11 4977 5009 4997FM 17 4983 5021 5003FF 16 5035 5038 5038MM 6 5060 4978 5027

General Summary and Conclusions

We have described the findings of two studies exploring the limits of remotemental influence effects We wished to see whether the mental strategiesadopted by the influencee would affect remote influence scores In particularcould an influencee block or resist a remote influence attempt We recorded in-fluenceesrsquo electrodermal activity while the influencer attempted to calm to ac-tivate or to have no influence over the influenceesrsquo physiology In each studythere were two conditions reflecting the influenceesrsquo designated mental strate-gies in response to the remote influence attempt blocking and co-operating

There were only two formal hypotheses 1) that remote influence scoreswould be significantly higher during the co-operating condition than duringthe blocking condition and 2) that overall there would be significant evidenceof remote mental influence on physiology in the designated direction (iecalming when calming was the influencerrsquos intention activating when activat-ing was the influencerrsquos intention) Contrary to prediction Study 1 found thatremote influence effects were higher in the block condition than in the co-op-erate condition (but not to a significant extent) Also overall remote influenceperformance was at chance level This study therefore found no support for ei-ther of the formal hypotheses Study 2 consisted of a replication of Study 1with a few minor adjustments plus additional measures that were not reportedin this paper Again contrary to prediction remote influence scores were non-significantly higher in the block condition than in the co-operate conditionThis replicates the trend found in Study 1 failing to support hypothesis oneThese results suggest that the mental strategy adopted by the influencee in theface of a remote influence attempt has no effect on remote influence perfor-mance Referring back to the questions posed in the introduction then wehave not found any evidence to suggest that it is possible to resist a remote in-fluence effect and there is as yet no indication of the possible limiting condi-tions for remote mental influence

Unlike Study 1 Study 2 found statistically significant evidence of an overallremote influence effect in the prespecified direction (ie calming of EDAwhen that was the influencerrsquos intention activation of EDA when that was theinfluencerrsquos intention) In order to facilitate comparison with a recent meta-an-alytic review of remote influence studies we calculated the effect size (r) ac-cording to the formula r = [t2(t2+df)]12 For their review of 19 EDA influencestudies Schlitz and Braud (1997) found an average effect size of r = +25 OurStudy 1 with overall t = - 031 df = 31 gives a negligible effect size of r = - 006 Study 2 with overall t = 1806 df = 49 gives r = +25 Study 2 there-fore replicates the mean effect size found in EDA studies conducted by otherresearchers in other parapsychology laboratories Post hoc we can combinethe data of the two studies for greater statistical power Overall t = 1176 df = 81 giving a small effect size of r = +13

We can find no clear reason why Study 2 found an EDA influence effectwhile Study 1 did not A number of factors are worth considering 1) The

Blocking vs Co-operating 531

532 C Watt et al

significant result could be due to chance or undetected artifact 2) Study 2 im-plemented what we considered to be minor methodological improvements thatmight have removed some artifactual fluctuations from the EDA data that weobserved in study one 3) Each experimenter in Study 2 conducted three testingsessions with their participant By the third session in which the remote influ-ence measure was taken experimenters were on quite friendly terms with theirparticipants perhaps more so than in Study 1 where each participant came infor only one testing session 4) The participants in Study 2 were more represen-tative of the general population we felt than those in Study 1 Study 1 in-volved more undergraduate psychology students and associates of the experi-menters 5) The experimenters in study one were less experienced with theremote influence protocol than when they conducted Study 2

In addition to the formal hypotheses we reported a number of exploratorymeasures Study 1 found that influenceesrsquo self-reported vividness of visual im-agery appeared to be unrelated to the remote influence performance In bothstudies influenceesrsquo qualitative descriptions of the actual mental strategiesthey employed when blocking and co-operating revealed a tendency for differ-ent strategies to be preferred for the different conditions The use of visual im-agery was popular for both blocking (eg visualizing being surrounded by aprotective cocoon) and co-operating (eg visualizing a river flowing betweeninfluencer and influencee) However a considerable proportion of participantsreported adopting an open or relaxed frame of mind when co-operating whilstalmost no participants chose this strategy when blocking This was despite thefact that prior to each session we encouraged participants to adopt comparablestrategies for blocking and co-operating

It was interesting and perhaps surprising to note that in both studies themajority of participants reported no feelings of concern or discomfort at thepossibility of having their physiology remotely influenced Those who wereconcerned tended to be those who knew more about parapsychology and thepossible harmful applications of remote influence A small proportion of par-ticipants was unconcerned because they did not believe such an influence waspossible in the first place

Both studies employed an exploratory variance measure of EDA influenceThis measure showed a greater range of scoring overall thus perhaps greaterpotential sensitivity to a remote influence effect However the variance mea-sure showed no indication of a remote influence effect To date researchershave employed a variety of indicators of remote influence as the literaturegrows progress may be made on identifying which is the most appropriate in-dicator

Neither study found significant sex differences in remote influence effectsthough in both studies female influencees tended to have higher remote influ-ence scores than male influencees

In terms of their overall remote influence effects there appeared to be nodifference between experimenterinfluencers in Study 1 (r = +06 - 05 and

- 06) In Study 2 effect size measures indicated that the experimenters dif-fered in remote influence effect sizes (r = +17 and +30) Due to the smaller Nthis difference was not found to be statistically significant Post hoc combin-ing Studies 1 and 2 for experimenters JW and CW (ZMcD was not an experi-menter in Study 2) the effect sizes are r = +08 and r = +20 respectively

We conducted a post hoc analysis of the effects of influenceeinfluencer sexpairings on remote influence scores Our two studies came up with contradic-tory findings each statistically significant Study one found that mixed sexpairings had significantly higher remote influence scores than same sex pair-ings This replicated the post hoc finding of Schlitz and LaBerge (1997) in con-nection with the autonomic detection of remote observation Study 2 foundthat same sex pairings had significantly higher remote influence scores thanmixed sex pairings Our contradictory findings do not help to clarify the pic-ture on the importance or otherwise of sex pairings It is quite possible that sexis not the most salient aspect of the dynamic between pairs Age personalityand familiarity are likely to be important variables moderating the relationshipbetween influencer and influencee

There are many more variables that we could look at but as the number ofanalyses multiplies so does the opportunity for spurious significant resultsFor this reason we have named only two formal hypotheses for each experi-ment All other analyses are regarded as exploratory and any trends that wehave highlighted above would require replication

In conclusion these two studies have found no evidence of a difference inremote influence effects when influencees are asked to adopt cooperative men-tal strategies compared to when they are asked to adopt blocking mental strate-gies In fact contrary to expectation both studies found slightly higher remoteinfluence effects for the blocking condition compared to the co-operating con-dition On this basis we have found no indication of limiting conditions for re-mote influence effects Therefore while anecdotal advice exists on how to de-fend oneself against a remote physiological influence attempt in the realworld laboratory evidence to date has provided no support for this advice

Given the importance of this question both in practical terms and in theoret-ical terms we argue that this line of research deserves continued investigationIf a way could be found to conduct research with a more realistic scenario mdash where the remote influence was not just neutrally and abstractlydefined as unwelcome mdash then our research might have more meaningful thingsto say about the parameters of remote influence It is interesting and perhapssurprising to note that when Watt has lectured on this topic to informed lay au-diences the feedback has often been that researchers must use more potential-ly harmful or disruptive remote influence (of course with the informed consentof the participants) While there are clear ethical inhibitions when consideringpossible physiological influence perhaps research using higher level influencewould be more acceptable For example one could attempt to disrupt or en-hance the concentration of chess players who were aware of the influence

Blocking vs Co-operating 533

534 C Watt et al

attempt and who were concurrently blocking or co-operating with the influ-ence attempt Research already exists to suggest that remote influence does notjust operate at physiological levels Attention focusing for example can be fa-cilitated by a remote helper (Brady amp Morris 1997 Braud et al 1995) Futureresearch in this area must therefore expand beyond physiological influences

Acknowledgements

The authors gratefully acknowledge the financial support they receivedfrom the Institut fuumlr Grenzgebiete der Psychologie und PsychohygieneFreiburg Germany while conducting these studies We would also like tothank Dr Paul Stevens for generously helping to solve software difficultieswhen preparing for these studies and the anonymous reviewer for helpfulcomments Address reprint requests to Dr C Watt Department of PsychologyUniversity of Edinburgh 7 George Square Edinburgh EH8 9JZ Scotland UK

ReferencesBrady C amp Morris R L (1997) Attention focusing facilitated through remote mental interac-

tion A replication and exploration of parameters In R Wiseman (Ed) Proceedings of Pre-sented Papers The Parapsychological Association 40th Annual Convention held in conjunc-tion with The Society for Psychical Research pp 73ndash91

Braud W G (1985) Blockingshielding psychic functioning through psychological and psychictechniques A report of three preliminary studies In RA White and J Solfvin (Eds) Re-search in Parapsychology 1984 Metuchen NJ Scarecrow Press pp 42ndash44

Braud W G amp Schlitz M J (1991) Consciousness interactions with remote biological systemsAnomalous intentionality effects Subtle Energies 2 1

Braud W G Schlitz M J Collins J and Klitch H (1985) Further studies of the bio-PK effectFeedback blocking specificitygenerality In RA White amp J Solfvin (Eds) Research inParapsychology 1984 Metuchen NJ Scarecrow Press pp 45ndash48

Braud W G Shafer D McNeill K and Guerra V (1995) Attention focusing facilitatedthrough remote mental interaction Journal of the American Society for Psychical Research89 103

Cohen J (1997) Statistical Power Analysis for the Behavioral Sciences New York AcademicPress

Dalton K (1994) Remotely influenced ESP performance in a computer task A preliminarystudy Proceedings of the 37th Annual Convention of the Parapsychological Association pp95ndash103

Dossey L (1994) Healing and the mind Is there a dark side Journal of Scientific Exploration 873

Halifax-Grof J (1974) Hex death In A Angoff amp D Barth (Eds) Parapsychology and Anthro-pology Proceedings of an International Conference Held in London England August 29ndash311973 NY Parapsychology Foundation

Hope M (1983) Practical Techniques of Psychic Self-Defence Wellingborough Northampton-shire Aquarian Press

Marks DF (1973) Visual imagery differences in the recall of pictures British Journal of Psy-chology 64 17

Nash CB (1982) Psychokinetic control of bacterial growth Journal of the Society for PsychicalResearch 51 217

Radin DI Taylor RK and Braud WG (1995) Remote mental influence of human electroder-mal activity A pilot replication European Journal of Parapsychology 11 19

Schlitz M amp Braud W G (1997) Distant intentionality and healing Assessing the evidence Al-ternative Therapies 3 62

Schlitz M J amp LaBerge S (1997) Covert observation increases skin conductance in subjectsunaware of when they are being observed A replication Journal of Parapsychology 61 185

Schmeidler G R (1958) Agent-percipient relationships Journal of the American Society forPsychical Research 52 47

Schmeidler G R (1961) Evidence for two kinds of telepathy International Journal of Parapsy-chology 3 5

Tart C T (1979) A survey of expert opinion on potentially negative uses of psi United Statesgovernment interest in psi and the level of research funding of the field In WG Roll (Ed) Re-search in Parapsychology 1978 Metuchen NJ Scarecrow Press pp 54

Tart C T amp Labore C M (1986) Attitudes towards strongly functioning psi A preliminary sur-vey Journal of the America Society for Psychical Research 80 163

Tart C T (1986) Psychicsrsquo fears of psychic powers Journal of the American Society for Psychi-cal Research 80 278

Watt C A McDermott Z amp Ravenscroft J (1997) Exploring the limits of remote mental inter-action A comparison of blocking versus co-operating strategies In R Wiseman (Ed) Pro-ceedings of the 40th Annual Convention of the Parapsychological Association held in conjunc-tion with the Society for Psychical Research Brighton England pp 464ndash476

Blocking vs Co-operating 535

520 C Watt et al

(ldquohey block nowrdquo ldquohey co-operate nowrdquo and ldquosession over thank yourdquo) tothe influencee just prior to each of the two 10-minute runs duplicating thewritten instructions that appeared on the computer monitor The physiologysensors that were attached to the influenceersquos fingers were located in the influ-enceersquos room The signals from the sensors were relayed along a cable andprocessed by a computer located in the influencerrsquos room Therefore the com-puter in the influenceersquos room was merely used to provide instructions and avisual display to the influencee The computer in the influencerrsquos room con-trolled the DMILS software and processes and stored the psychophysiologydata

Procedure

The influencer and participant initially sat down in the reception room in theparapsychology suite for a further description of procedure and returned theircompleted PIFimagery questionnaire to the influenceer who went over it withthe participant Any questions that were raised by the participant prior to test-ing were answered here (This question-and-answer session usually lasted fif-teen minutes) The participant then gave their informed consent to the studyThe influencer also suggested various mental approaches the participant mightadopt for the blocking and co-operating conditions (some examples weregiven in the information sheet) These suggestions included both visual im-agery (eg visualizing a shielding cocoon in the blocking condition) and cog-nitive strategies (eg adopting a stubborn and uncooperative frame of mind inthe blocking condition) Participants were informed that different strategiesmight suit different people and they were to do whatever felt natural for themThe influencer then took the participant to the testing room and demonstratedthe auditory instructions so that the participant would not be startled by thesewhen they occurred during the session itself The participant was also taken tosee where the influencer was sitting and the influencer described what theywould be doing when they were attempting to activate or calm the participant

Once the preliminaries were finished the influencee was seated in a partiallyreclined chair in front of the computer monitor and the physiology sensorswere attached to their fingers by the influencer The room was softly illuminat-ed by a desk lamp with an orangered filter At this stage the influencer checkedthat the physiology signals were being transmitted to the computer in theirroom Returning to the influencee the influencer set a start time on the influ-enceersquos computer This was on average two to three minutes later The influ-encer then returned to their room closing all doors and waited for the start ofthe session

Physiology Data Collection As the session began the influencee was in-structed visually and auditorially either to ldquohey block nowrdquo or ldquohey co-oper-ate nowrdquo The influenceersquos monitor then displayed a colorful and dynamicscreen-saver that was intended to be engaging without being particularlyarousing or relaxing At the same time the influencer was following the influ-

ence schedule as instructed by the computer observing the real-time feedbackdisplay of the participantrsquos EDA when desired At the end of the first run whichlasted approximately ten minutes the influencer saw a 150 second count downon the computer before the second run began At the same time the influenceesaw a message about the strategy to adopt for the forthcoming second run (ieto block or to co-operate whichever strategy was not used for the first run) Atthe beginning of the second run the influencee was also instructed auditoriallyeither to ldquohey block nowrdquo or ldquohey co-operate nowrdquo The dynamic visual dis-play resumed while the influencer followed a random influence schedule forapproximately ten minutes The entire session took about 25 minutes to com-plete At the end of the session the influencee was auditorially and visually in-formed ldquosession over thank yourdquo and waited until the influencer returned tothe influenceersquos room

Post-Testing Procedures The influenceersquos physiology sensors were re-moved by the influencer and both returned to the reception room to completethe post-session questionnaires Then the influencee was taken to the influ-encerrsquos room to see feedback on the results of their testing session There wasan option for some final discussion about the session after which the volun-teer was thanked for their participation and left Once all the data was collect-ed feedback on the results of the study as a whole was sent to participants

Planned Hypotheses and Analyses

Our main aim was to examine whether there was any evidence for blockingof a remote mental influence on physiology [EDA is the principal physiologymeasure here with per-session PSI scores representing the measure of theDMILS effect on EDA] H1 therefore predicted significantly greater PSIscores in the co-operating condition than in the blocking condition (using a re-lated t-test one-tailed) Our second question was whether there was any over-all evidence of remote mental influence on EDA This analysis would be con-ducted on the data of blocking and co-operating runs combined if nosignificant difference was found between these two conditions (as tested byH1) If H1 was supported H2 would be tested on the data of co-operating runsonly predicting PSI scores significantly greater than chance expectation(using a single-mean t-test of the PSI score vs MCE one-tailed)

Planned alpha level was 05 There was no correction for multiple analysisThis was dealt with by restricting the number of formal hypotheses (with one-tailed significance tests) to two in each study All other analyses were regardedas exploratory and two-tailed significance tests were used

Exploratory Measures

Variance

DMILS researchers do not yet appear to have identified what might be themost appropriate indicator of a remote influence effect Therefore in addition

Blocking vs Co-operating 521

522 C Watt et al

to the PSI score we looked at the amount of variance that was shown in block-ing and co-operating scores The variance measure was calculated as followsMean values for activate periods and calm periods were calculated separatelyfor block and co-operate conditions Mean values were taken from each indi-vidual score in that condition and squared and summed S [mean value for con-dition - individual score]2 This gave a summation of variability around themean for each activate calm and rest period for block and cooperate condi-tions It was used in conjunction with the mean level of EDA activity givingan indication of the motion of the wave (ie the variability of the EDA activi-ty) rather than the overall level

VVIQ

Anecdotally the use of appropriate imagery (eg visualize being surroundedby a shielding white halo of white light) is suggested to be an effective blockingtechnique (eg Hope 1983) We therefore wished to see if participants who re-ported experiencing vivid imagery with ease are perhaps more effective block-ers than those who do not report clear mental imagery Descriptive data wasgathered to explore any trends in PSI scores for blockingco-operating forthose participants with VVIQ scores ranging from low (ie reporting vividmental imagery) to high (ie reporting little or no mental imagery)

Qualitative Reports

We also wanted to report data illustrating what kind of different strategiesparticipants used when they were blocking and co-operating These qualitativereports were gathered in the form of the participantsrsquo written responses toopen-ended questions on a post-session questionnaire Watt and Ravenscroftindependently coded these written responses into categories and cross-checked to ensure coding agreement

Sex and Experimenter Effects

We also examined the data for sex effects and experimenter effects becausethese variables could confound the interpretation of any overall effects Final-ly a post hoc analysis was conducted on influenceeinfluencer sex pairings Ina study on the unconscious detection of remote staring Schlitz and LaBerge(1997) found the greatest effect for opposite-sex observer-subject pairs

Study One Results and Discussion

Participants

Thirty-two individuals took part as influencees (12 females 20 males) theirages ranged from 19 to 63 years (mean age 285 years)

Incomplete Data

One participant declined to give their age on the PIF form so the descriptivestatistics on age do not include this participant (so N = 31) Also due to a tech-nical hitch the exploratory variance scores were lost for one participant so N = 31 for this measure There were no drop-outs from this study

Remote Influence Measure

Table 1 presents a summary of the remote influence data in the form ofmean Percentage Score Index for co-operating and blocking conditions Therewas no significant difference between PSI for co-operating and blocking con-ditions (t = - 202 df = 31) Therefore Hypothesis 1 (predicting significantlygreater PSI scores for co-operating condition compared to blocking condition)was not supported by these results As planned the PSI scores for co-operatingand blocking runs were combined to examine Hypothesis 2 (of an overall re-mote mental influence effect in line with the calming and activating intentionsof the influencee) Mean PSI scores were exactly at chance expectation (t = - 031 df = 31) therefore there was no support for Hypothesis 2

Exploratory Questions

Variance Measure

Table 2 presents summary data for the exploratory measure of the amount ofvariance shown in each condition (lt 50 represents reduced variance gt 50 rep-resents increased variance) There was no significant difference between vari-ance scores for co-operating and blocking conditions (t = 835 df = 30) andwhen the conditions were combined variance did not depart from chance ex-pectation (t = - 244 df = 30) It is interesting to note however that the vari-ance measure gives a much wider range of scores than the PSI measure

Blocking vs Co-operating 523

TABLE 1Summary Data Mean PSI Scores

(N = 32 MCE = 50 range in brackets)

Co-operate PSI Block PSI Combined PSI

4996 5003 5000(4692- 5215) (4653- 5404) (4853- 5309)

TABLE 2Summary Data Mean Variance Scores(N = 31 MCE = 50 range in brackets)

Co-operate Variance Block Variance Combined Variance

5089 4864 4977(2648- 7012) (2492- 6484) (4116- 6297)

524 C Watt et al

suggesting that the variance measure may potentially be a more sensitive indi-cator of remote influence effect than PSI

Vividness of Mental Imagery

Low VVIQ scores represent participants reporting highly vivid imagery andhigh VVIQ scores represent those reporting little or no imagery The meanscore was 4275 (range from 21 to 80) Using a Pearson correlation we foundthat there was no evidence to suggest that those participants with low VVIQscores (high imagery) were better blockers than those with high scores (VVIQvs PSI blocking r = - 113) We also found that there was no evidence of a cor-relation between vividness of imagery and overall remote influence effectsusing the combined Percentage Score Index measure (VVIQ vs PSI overall r =- 077)

Influenceesrsquo Blocking and Co-operating Strategies

Influencees completed a questionnaire (after the session was over) askingthem to describe what strategies they used for blocking and co-operating Cat-egorizing these responses we found a difference in the kinds of strategies usedwhen blocking compared to co-operating as shown in Table 3 In the blockingcondition the majority of participants (469) used visual imagery strategiesFewer (281) participants used a combination of imagery and cognitivestrategies (such as repeating words related to blocking) while 25 reportedusing purely cognitive strategies for blocking (such as adopting a resistant anduncooperative attitude or thinking of irrelevancies) In the co-operating con-dition fewer participants used visual imagery alone (333) The majority ofparticipants (424) reported that their strategies for co-operating includedadopting a feeling of passiveness openness or relaxation whereas no partici-pants adopted this kind of strategy for the blocking condition 121 of partic-ipants used cognitive strategies alone (such as adopting a co-operative atti-tude) and the same proportion used a combination of visual imagery andcognitive strategies

Therefore there was a tendency for participants to use visual imagery forblocking and relaxationpassivity or visual imagery for co-operating This wasdespite the fact that we asked participants to attempt to adopt similar levels ofactivation for both co-operating and blocking strategies There was also a ten-dency not surprisingly for those participants with low VVIQ scores (ie vivid

TABLE 3Influenceesrsquo Mental Strategies ()

Vis Im Cognitive VisCog OpenRelax

Blocking 469 250 281 00Co-operating 333 121 121 424

imagers) to use visual imagery strategies Those who had high VVIQ scoresalso used visual imagery but more often used this in combination with othertactics

Other Qualitative Reports from Participants According to their post-ses-sion questionnaire responses participants generally felt comfortable about theattempt to be influenced (N = 24) Three participants said they felt apprehen-sive or uncomfortable with the idea of being influenced Interestingly of thesethree two were actively involved with parapsychology and were aware of theharmful psi implications discussed at the beginning of this report The thirdwas a psychology student who had attended a lecture given by Watt in whichthe possibilities of harmful psi were discussed Another group of three partici-pants said they would be uncomfortable with the idea of being influenced ifthey believed such influence was possible

Seven participants said they found it was difficult to sustain concentrationfor the 10 minutes that comprised each condition Five participants found thescreensaver distracting or disturbing Two participants specifically mentionedthat they found blocking a more taxing task than co-operating

Sex Effects

There were more males (N = 20) than females (N = 12) in this study howev-er there was no significant difference in PSI scores for remote physiology in-fluence between males and females (male PSI vs female PSI t = 1085 male PSI = 4987 female PSI = 5021)

Experimenter Effects

We studied whether there was any difference in PSI scores obtained by thethree experimenters The results are shown in Table 4 There seems to be nosign of experimenter effects in study one

InfluenceeInfluencer Sex Pairings

Table 5 gives the mean PSI scores for different influenceeinfluencer pair-ings Taking the combined PSI scores it can be seen that the highest remote in-fluence effect is for FM pairs followed by MF pairs followed by FF pairsand the lowest scoring is for MM pairs Post hoc it was found that the mixedsex pairings showed significantly higher remote influence scores than the same

Blocking vs Co-operating 525

TABLE 4Remote Influence Results by Experimenter

Experimenter N sessions Coop PSI Block PSI Combined PSI

CW 12 4983 5011 4997ZMcD 10 5015 4998 5006JR 10 4994 4999 4996

526 C Watt et al

sex pairings (t = - 2016 df = 30 p = 053 2 - t) This replicates the effect foundby Schlitz and LaBerge (1997)

Study One Conclusions

In Study 1 we developed a methodology to examine intentional resistance toa remote mental influence attempt (Watt et al 1997) We contrasted influ-enceesrsquo physiological activity when they were mentally co-operating with aremote attempt to influence their physiology with physiological activity wheninfluencees were mentally blocking the influence attempt We found no evi-dence for any remote influence effect overall and no difference betweenblocking and co-operating conditions Our qualitative measures yielded someinteresting information about participantsrsquo mental strategies and reactions dur-ing the remote influence session We found no difference between experi-menters in remote influence scores but we did find that mixed sex influ-enceeinfluencer pairings scored more highly than same sex pairings

Study Two Method

Study 2 was planned to explore in more depth the nature of the relationshipbetween aspects of personality and cognition and psi performance Partici-pants took part in three testing sessions approximately one week apart Thefirst consisted of various psychological measures the second consisted of aforced choice ESP task and the third consisted of a replication of the blocking-co-operating study (Study 1) described above The psychological measureswere not scored until after completion of the remote influence session Partici-pants received feedback about their forced choice ESP performance at the con-clusion of the second session For ease of interpretation the present paper isrestricted to a description of the third testing session which was essentially areplication of study one with fewer influencers and greater statistical power

We introduced some slight changes to the methodology of Study 2 based onour experiences with Study 1 At the start of each of the two runs in Study 1the participant was given auditory instructions either to ldquoblock nowrdquo or ldquoco-operate nowrdquo We observed that the auditory instructions appeared to affect theparticipantrsquos EDA as if they were startled As planned the first recordingepoch in each run was not used in analysis to give the participantrsquos EDA time

TABLE 5Mean PSI Scores for Different InfluenceeInfluencer Sex Pairings

Pairing N Coop PSI Block PSI Combined PSI

M-F 6 4932 5069 5001F-M 6 5043 5034 5048F-F 6 5022 4952 4993M-M 14 4988 4975 4981

to settle down However in order to remove any remaining potential artifactu-al effects on EDA in Study 2 only visual instructions were given

Another difference in methodology concerned the screensaver viewed bythe participants during the experimental session In Study 1 some participantsreported that the series of patterns produced by the computer were being dis-played at too fast a rate such that some participants found it distracting or dis-turbing Again we felt that this may have an artifactual effect on EDA so wedecided for study two to reduce the speed of the screensaver

Thirdly in Study 2 we did not give participants detailed information abouttheir remote influence performance until the feedback letters that were distrib-uted at the conclusion of the entire study The reason for this change was mere-ly to streamline the procedure on what was already a time-consuming experi-ment Finally in another streamlining change we did not take VVIQ measuresin Study 2

Planned Hypotheses and Analyses

These were the same as with Study 1 That is hypothesis 1 predicted higherPSI scores for the co-operating condition than for the blocking condition(measured with a related t-test one-tailed) If no significant difference wasfound between blocking PSI and co-operating PSI scores the two would becombined to look at overall remote influence effect (hypothesis 2 single-meant-test one-tailed) If a significant difference was found with hypothesis 1 theanalysis for hypothesis 2 would be conducted with the data for co-operatingsessions only

Exploratory Questions

As with study one we took an exploratory variance measure of remote EDAinfluence We also gathered qualitative descriptions of the mental strategiesparticipants adopted whilst blocking and co-operating We also looked at sexeffects experimenter effects and post hoc influenceeinfluencer sex pairings

Participants

Study 2 had two experimenterinfluencers (Watt amp Ravenscroft) After aparticipant was recruited (usually by Ravenscroft) they were assigned to oneof the two experimenters for all three testing sessions This assignment wasdone unsystematically according to the preferences and availability of the ex-perimenters As planned 50 volunteer participants were drawn from the gener-al participant pool at the Koestler Parapsychology Unit and by word ofmouth All participants completed the Participant Information Form and aConsent form prior to the study The participants consisted of 33 females and17 males Average age was 41 years range from 22 years to 76 years of age Inorder to increase the studyrsquos relevance to the general population undergradu-ate psychology students were not recruited as participants The main criterion

Blocking vs Co-operating 527

528 C Watt et al

for participation was that the participants were interested in and open-mindedabout parapsychology Again there were no drop-outs from this study Allparticipants received a detailed personal feedback letter at the conclusion ofthe study

Study Two Results and Discussion

Remote Influence Measure

PSI scores are summarized in Table 6 which shows that the mean PSI scorewhen co-operating was nonsignificantly lower than mean PSI score whenblocking (t = - 595 df = 49) Hypothesis 1 was therefore not supported Asplanned in the absence of any significant difference between blocking and co-operating PSI scores the two conditions were combined to test hypothesis 2which predicted overall significant remote mental influence performance Thecombined PSI score was significantly greater than mean chance expectation of50 (t = 1806 df = 49 p = 038 1 - t) Therefore hypothesis 2 was supportedgiving evidence of an overall remote influence effect upon psychophysiologyin the prespecified direction

The failure to support hypothesis 1 confirms what we found in Study 1 thatthere was no significant difference between blocking and co-operating condi-tions on remote influence scores Indeed both studies found a slight effect inthe opposite direction to that predicted that is slightly higher PSI scores whenblocking was the influenceersquos intention than when co-operating was their in-tention It was encouraging to find significant support for a remote influenceeffect in the predicted direction for blocking and co-operating conditions com-bined No such trend was found in the previous study though with a smallernumber of participants (N = 32) the earlier study had lower statistical powerthan the present study

Exploratory Questions

Variance Measure

Table 7 shows the results for the exploratory measure of EDA variance lt 50represents reduced variance gt 50 represents increased variance We found nodifference between variance scores for cooperate versus block conditions(t = 150 df = 49) When these are combined the resulting figure does not de-viate from chance expectation (t = 440 df = 49) Thus as in Study 1 it seems

TABLE 6Summary Data Mean PSI Scores

(N = 50 MCE = 50 range in brackets)

Co-operate PSI Block PSI Combined PSI

5008 5019 5016(4821- 5221) (4807- 5557) (4900- 5296)

that the exploratory variance measure does not reveal any additional indica-tion of the operation of a remote mental influence in this study However thewider range of variance scoring continues to suggest that this measure mayhave greater potential sensitivity than the summary PSI score

Influenceesrsquo Blocking and Co-operating Strategies

We were interested in learning more about what particular strategies partici-pants chose to adopt when blocking and co-operating As in Study 1 a post-session questionnaire was used to gather this information Table 8 shows thepercentage of participants who adopted each kind of strategy

The table shows a very similar pattern to that found in Study 1 The mainpoint of interest is that almost no participants took a passive approach toblocking (ie adopting an openrelaxing strategy) A much higher proportionpreferred to adopt this approach when co-operating The preferred mode forboth conditions was to use some form of visual imagery such as imagining ashield or a screen when blocking and imagining an open channel such as ariver when co-operating

The majority of participants reported no concerns over the possibility ofhaving their physiology remotely influenced by another person as found instudy one

Sex Effects

As in study one the sexes were not equally represented among participantsWhereas Study 1 had an excess of male participants in Study 2 twice as manyfemales (N = 33) as males (N = 17) participated It was therefore important tolook for sex differences since any such differences could confound the inter-pretation of overall trends We found no difference between sexes when co-op-erating (female mean PSI score = 5008 male mean PSI score = 5006 t = 107 df = 48) There was a nonsignificant trend for females to have higher

Blocking vs Co-operating 529

TABLE 7Summary Data Mean Variance Scores(N = 50 MCE = 50 range in brackets)

Co-operate Variance Block Variance Combined Variance

5076 5054 5034(3164- 7279) (3572- 6813) (3749- 6671)

TABLE 8Influenceesrsquo Mental Strategies ()

Visual Imagery Cognitive VisCog OpenRelax

Blocking 64 26 6 2Co-operating 60 8 12 20

530 C Watt et al

PSI scores than males when blocking (mean PSI scores 5030 and 4998 re-spectively t = 992 df = 48) Combining blocking and co-operating conditionsthere was no overall difference between the sexes (mean PSI scores 5020 forfemales 5008 for males t = 665 df = 48)

Experimenter Effects

Study 2 found a statistically significant overall remote influence effect(combined PSI = 50155 t = 1806) Table 9 shows the breakdown of remoteinfluence scores for the two experimenters CW and JR Looking at the overalldata for the two experimenters we see that CWrsquos sessions were associatedwith most of the remote influence effect However due to the smaller N CWrsquossessions are not independently significant (t = 1632) and there is not a signifi-cant difference between CW and JRrsquos overall PSI scores (t = 678) Post hocwe calculated effect size r according to the formula r = [t2(t2+df)]12 for the twoexperimenters For JR (t = 817 df = 22) we found a small effect sizer = +17 For CW (t = 1632 df = 26) the effect size was moderate r = +30This may be some post hoc indication of an experimenter effect in Study 2

InfluenceeInfluencer Sex Pairings

As can be seen from Table 10 Study 2 found exactly the opposite trend tothat found in Study 1 On a post hoc test same sex pairings scored significant-ly higher on the remote influence PSI measure than mixed sex pairings(t = - 2034 df = 48 p = 048 2 - t) This contradicts Schlitz and LaBergersquos(1997) finding of greater autonomic detection of remote staring for mixed sexpairs

TABLE 9Remote Influence Results by Experimenter

Experimenter N sessions Coop PSI Block PSI Combined PSI

CW 27 5011 5026 5021JR 23 5003 5010 5009

TABLE 10Mean PSI Scores for Different InfluenceeInfluencer Sex Pairings

Pairing N Coop PSI Block PSI Combined PSI

MF 11 4977 5009 4997FM 17 4983 5021 5003FF 16 5035 5038 5038MM 6 5060 4978 5027

General Summary and Conclusions

We have described the findings of two studies exploring the limits of remotemental influence effects We wished to see whether the mental strategiesadopted by the influencee would affect remote influence scores In particularcould an influencee block or resist a remote influence attempt We recorded in-fluenceesrsquo electrodermal activity while the influencer attempted to calm to ac-tivate or to have no influence over the influenceesrsquo physiology In each studythere were two conditions reflecting the influenceesrsquo designated mental strate-gies in response to the remote influence attempt blocking and co-operating

There were only two formal hypotheses 1) that remote influence scoreswould be significantly higher during the co-operating condition than duringthe blocking condition and 2) that overall there would be significant evidenceof remote mental influence on physiology in the designated direction (iecalming when calming was the influencerrsquos intention activating when activat-ing was the influencerrsquos intention) Contrary to prediction Study 1 found thatremote influence effects were higher in the block condition than in the co-op-erate condition (but not to a significant extent) Also overall remote influenceperformance was at chance level This study therefore found no support for ei-ther of the formal hypotheses Study 2 consisted of a replication of Study 1with a few minor adjustments plus additional measures that were not reportedin this paper Again contrary to prediction remote influence scores were non-significantly higher in the block condition than in the co-operate conditionThis replicates the trend found in Study 1 failing to support hypothesis oneThese results suggest that the mental strategy adopted by the influencee in theface of a remote influence attempt has no effect on remote influence perfor-mance Referring back to the questions posed in the introduction then wehave not found any evidence to suggest that it is possible to resist a remote in-fluence effect and there is as yet no indication of the possible limiting condi-tions for remote mental influence

Unlike Study 1 Study 2 found statistically significant evidence of an overallremote influence effect in the prespecified direction (ie calming of EDAwhen that was the influencerrsquos intention activation of EDA when that was theinfluencerrsquos intention) In order to facilitate comparison with a recent meta-an-alytic review of remote influence studies we calculated the effect size (r) ac-cording to the formula r = [t2(t2+df)]12 For their review of 19 EDA influencestudies Schlitz and Braud (1997) found an average effect size of r = +25 OurStudy 1 with overall t = - 031 df = 31 gives a negligible effect size of r = - 006 Study 2 with overall t = 1806 df = 49 gives r = +25 Study 2 there-fore replicates the mean effect size found in EDA studies conducted by otherresearchers in other parapsychology laboratories Post hoc we can combinethe data of the two studies for greater statistical power Overall t = 1176 df = 81 giving a small effect size of r = +13

We can find no clear reason why Study 2 found an EDA influence effectwhile Study 1 did not A number of factors are worth considering 1) The

Blocking vs Co-operating 531

532 C Watt et al

significant result could be due to chance or undetected artifact 2) Study 2 im-plemented what we considered to be minor methodological improvements thatmight have removed some artifactual fluctuations from the EDA data that weobserved in study one 3) Each experimenter in Study 2 conducted three testingsessions with their participant By the third session in which the remote influ-ence measure was taken experimenters were on quite friendly terms with theirparticipants perhaps more so than in Study 1 where each participant came infor only one testing session 4) The participants in Study 2 were more represen-tative of the general population we felt than those in Study 1 Study 1 in-volved more undergraduate psychology students and associates of the experi-menters 5) The experimenters in study one were less experienced with theremote influence protocol than when they conducted Study 2

In addition to the formal hypotheses we reported a number of exploratorymeasures Study 1 found that influenceesrsquo self-reported vividness of visual im-agery appeared to be unrelated to the remote influence performance In bothstudies influenceesrsquo qualitative descriptions of the actual mental strategiesthey employed when blocking and co-operating revealed a tendency for differ-ent strategies to be preferred for the different conditions The use of visual im-agery was popular for both blocking (eg visualizing being surrounded by aprotective cocoon) and co-operating (eg visualizing a river flowing betweeninfluencer and influencee) However a considerable proportion of participantsreported adopting an open or relaxed frame of mind when co-operating whilstalmost no participants chose this strategy when blocking This was despite thefact that prior to each session we encouraged participants to adopt comparablestrategies for blocking and co-operating

It was interesting and perhaps surprising to note that in both studies themajority of participants reported no feelings of concern or discomfort at thepossibility of having their physiology remotely influenced Those who wereconcerned tended to be those who knew more about parapsychology and thepossible harmful applications of remote influence A small proportion of par-ticipants was unconcerned because they did not believe such an influence waspossible in the first place

Both studies employed an exploratory variance measure of EDA influenceThis measure showed a greater range of scoring overall thus perhaps greaterpotential sensitivity to a remote influence effect However the variance mea-sure showed no indication of a remote influence effect To date researchershave employed a variety of indicators of remote influence as the literaturegrows progress may be made on identifying which is the most appropriate in-dicator

Neither study found significant sex differences in remote influence effectsthough in both studies female influencees tended to have higher remote influ-ence scores than male influencees

In terms of their overall remote influence effects there appeared to be nodifference between experimenterinfluencers in Study 1 (r = +06 - 05 and

- 06) In Study 2 effect size measures indicated that the experimenters dif-fered in remote influence effect sizes (r = +17 and +30) Due to the smaller Nthis difference was not found to be statistically significant Post hoc combin-ing Studies 1 and 2 for experimenters JW and CW (ZMcD was not an experi-menter in Study 2) the effect sizes are r = +08 and r = +20 respectively

We conducted a post hoc analysis of the effects of influenceeinfluencer sexpairings on remote influence scores Our two studies came up with contradic-tory findings each statistically significant Study one found that mixed sexpairings had significantly higher remote influence scores than same sex pair-ings This replicated the post hoc finding of Schlitz and LaBerge (1997) in con-nection with the autonomic detection of remote observation Study 2 foundthat same sex pairings had significantly higher remote influence scores thanmixed sex pairings Our contradictory findings do not help to clarify the pic-ture on the importance or otherwise of sex pairings It is quite possible that sexis not the most salient aspect of the dynamic between pairs Age personalityand familiarity are likely to be important variables moderating the relationshipbetween influencer and influencee

There are many more variables that we could look at but as the number ofanalyses multiplies so does the opportunity for spurious significant resultsFor this reason we have named only two formal hypotheses for each experi-ment All other analyses are regarded as exploratory and any trends that wehave highlighted above would require replication

In conclusion these two studies have found no evidence of a difference inremote influence effects when influencees are asked to adopt cooperative men-tal strategies compared to when they are asked to adopt blocking mental strate-gies In fact contrary to expectation both studies found slightly higher remoteinfluence effects for the blocking condition compared to the co-operating con-dition On this basis we have found no indication of limiting conditions for re-mote influence effects Therefore while anecdotal advice exists on how to de-fend oneself against a remote physiological influence attempt in the realworld laboratory evidence to date has provided no support for this advice

Given the importance of this question both in practical terms and in theoret-ical terms we argue that this line of research deserves continued investigationIf a way could be found to conduct research with a more realistic scenario mdash where the remote influence was not just neutrally and abstractlydefined as unwelcome mdash then our research might have more meaningful thingsto say about the parameters of remote influence It is interesting and perhapssurprising to note that when Watt has lectured on this topic to informed lay au-diences the feedback has often been that researchers must use more potential-ly harmful or disruptive remote influence (of course with the informed consentof the participants) While there are clear ethical inhibitions when consideringpossible physiological influence perhaps research using higher level influencewould be more acceptable For example one could attempt to disrupt or en-hance the concentration of chess players who were aware of the influence

Blocking vs Co-operating 533

534 C Watt et al

attempt and who were concurrently blocking or co-operating with the influ-ence attempt Research already exists to suggest that remote influence does notjust operate at physiological levels Attention focusing for example can be fa-cilitated by a remote helper (Brady amp Morris 1997 Braud et al 1995) Futureresearch in this area must therefore expand beyond physiological influences

Acknowledgements

The authors gratefully acknowledge the financial support they receivedfrom the Institut fuumlr Grenzgebiete der Psychologie und PsychohygieneFreiburg Germany while conducting these studies We would also like tothank Dr Paul Stevens for generously helping to solve software difficultieswhen preparing for these studies and the anonymous reviewer for helpfulcomments Address reprint requests to Dr C Watt Department of PsychologyUniversity of Edinburgh 7 George Square Edinburgh EH8 9JZ Scotland UK

ReferencesBrady C amp Morris R L (1997) Attention focusing facilitated through remote mental interac-

tion A replication and exploration of parameters In R Wiseman (Ed) Proceedings of Pre-sented Papers The Parapsychological Association 40th Annual Convention held in conjunc-tion with The Society for Psychical Research pp 73ndash91

Braud W G (1985) Blockingshielding psychic functioning through psychological and psychictechniques A report of three preliminary studies In RA White and J Solfvin (Eds) Re-search in Parapsychology 1984 Metuchen NJ Scarecrow Press pp 42ndash44

Braud W G amp Schlitz M J (1991) Consciousness interactions with remote biological systemsAnomalous intentionality effects Subtle Energies 2 1

Braud W G Schlitz M J Collins J and Klitch H (1985) Further studies of the bio-PK effectFeedback blocking specificitygenerality In RA White amp J Solfvin (Eds) Research inParapsychology 1984 Metuchen NJ Scarecrow Press pp 45ndash48

Braud W G Shafer D McNeill K and Guerra V (1995) Attention focusing facilitatedthrough remote mental interaction Journal of the American Society for Psychical Research89 103

Cohen J (1997) Statistical Power Analysis for the Behavioral Sciences New York AcademicPress

Dalton K (1994) Remotely influenced ESP performance in a computer task A preliminarystudy Proceedings of the 37th Annual Convention of the Parapsychological Association pp95ndash103

Dossey L (1994) Healing and the mind Is there a dark side Journal of Scientific Exploration 873

Halifax-Grof J (1974) Hex death In A Angoff amp D Barth (Eds) Parapsychology and Anthro-pology Proceedings of an International Conference Held in London England August 29ndash311973 NY Parapsychology Foundation

Hope M (1983) Practical Techniques of Psychic Self-Defence Wellingborough Northampton-shire Aquarian Press

Marks DF (1973) Visual imagery differences in the recall of pictures British Journal of Psy-chology 64 17

Nash CB (1982) Psychokinetic control of bacterial growth Journal of the Society for PsychicalResearch 51 217

Radin DI Taylor RK and Braud WG (1995) Remote mental influence of human electroder-mal activity A pilot replication European Journal of Parapsychology 11 19

Schlitz M amp Braud W G (1997) Distant intentionality and healing Assessing the evidence Al-ternative Therapies 3 62

Schlitz M J amp LaBerge S (1997) Covert observation increases skin conductance in subjectsunaware of when they are being observed A replication Journal of Parapsychology 61 185

Schmeidler G R (1958) Agent-percipient relationships Journal of the American Society forPsychical Research 52 47

Schmeidler G R (1961) Evidence for two kinds of telepathy International Journal of Parapsy-chology 3 5

Tart C T (1979) A survey of expert opinion on potentially negative uses of psi United Statesgovernment interest in psi and the level of research funding of the field In WG Roll (Ed) Re-search in Parapsychology 1978 Metuchen NJ Scarecrow Press pp 54

Tart C T amp Labore C M (1986) Attitudes towards strongly functioning psi A preliminary sur-vey Journal of the America Society for Psychical Research 80 163

Tart C T (1986) Psychicsrsquo fears of psychic powers Journal of the American Society for Psychi-cal Research 80 278

Watt C A McDermott Z amp Ravenscroft J (1997) Exploring the limits of remote mental inter-action A comparison of blocking versus co-operating strategies In R Wiseman (Ed) Pro-ceedings of the 40th Annual Convention of the Parapsychological Association held in conjunc-tion with the Society for Psychical Research Brighton England pp 464ndash476

Blocking vs Co-operating 535

ence schedule as instructed by the computer observing the real-time feedbackdisplay of the participantrsquos EDA when desired At the end of the first run whichlasted approximately ten minutes the influencer saw a 150 second count downon the computer before the second run began At the same time the influenceesaw a message about the strategy to adopt for the forthcoming second run (ieto block or to co-operate whichever strategy was not used for the first run) Atthe beginning of the second run the influencee was also instructed auditoriallyeither to ldquohey block nowrdquo or ldquohey co-operate nowrdquo The dynamic visual dis-play resumed while the influencer followed a random influence schedule forapproximately ten minutes The entire session took about 25 minutes to com-plete At the end of the session the influencee was auditorially and visually in-formed ldquosession over thank yourdquo and waited until the influencer returned tothe influenceersquos room

Post-Testing Procedures The influenceersquos physiology sensors were re-moved by the influencer and both returned to the reception room to completethe post-session questionnaires Then the influencee was taken to the influ-encerrsquos room to see feedback on the results of their testing session There wasan option for some final discussion about the session after which the volun-teer was thanked for their participation and left Once all the data was collect-ed feedback on the results of the study as a whole was sent to participants

Planned Hypotheses and Analyses

Our main aim was to examine whether there was any evidence for blockingof a remote mental influence on physiology [EDA is the principal physiologymeasure here with per-session PSI scores representing the measure of theDMILS effect on EDA] H1 therefore predicted significantly greater PSIscores in the co-operating condition than in the blocking condition (using a re-lated t-test one-tailed) Our second question was whether there was any over-all evidence of remote mental influence on EDA This analysis would be con-ducted on the data of blocking and co-operating runs combined if nosignificant difference was found between these two conditions (as tested byH1) If H1 was supported H2 would be tested on the data of co-operating runsonly predicting PSI scores significantly greater than chance expectation(using a single-mean t-test of the PSI score vs MCE one-tailed)

Planned alpha level was 05 There was no correction for multiple analysisThis was dealt with by restricting the number of formal hypotheses (with one-tailed significance tests) to two in each study All other analyses were regardedas exploratory and two-tailed significance tests were used

Exploratory Measures

Variance

DMILS researchers do not yet appear to have identified what might be themost appropriate indicator of a remote influence effect Therefore in addition

Blocking vs Co-operating 521

522 C Watt et al

to the PSI score we looked at the amount of variance that was shown in block-ing and co-operating scores The variance measure was calculated as followsMean values for activate periods and calm periods were calculated separatelyfor block and co-operate conditions Mean values were taken from each indi-vidual score in that condition and squared and summed S [mean value for con-dition - individual score]2 This gave a summation of variability around themean for each activate calm and rest period for block and cooperate condi-tions It was used in conjunction with the mean level of EDA activity givingan indication of the motion of the wave (ie the variability of the EDA activi-ty) rather than the overall level

VVIQ

Anecdotally the use of appropriate imagery (eg visualize being surroundedby a shielding white halo of white light) is suggested to be an effective blockingtechnique (eg Hope 1983) We therefore wished to see if participants who re-ported experiencing vivid imagery with ease are perhaps more effective block-ers than those who do not report clear mental imagery Descriptive data wasgathered to explore any trends in PSI scores for blockingco-operating forthose participants with VVIQ scores ranging from low (ie reporting vividmental imagery) to high (ie reporting little or no mental imagery)

Qualitative Reports

We also wanted to report data illustrating what kind of different strategiesparticipants used when they were blocking and co-operating These qualitativereports were gathered in the form of the participantsrsquo written responses toopen-ended questions on a post-session questionnaire Watt and Ravenscroftindependently coded these written responses into categories and cross-checked to ensure coding agreement

Sex and Experimenter Effects

We also examined the data for sex effects and experimenter effects becausethese variables could confound the interpretation of any overall effects Final-ly a post hoc analysis was conducted on influenceeinfluencer sex pairings Ina study on the unconscious detection of remote staring Schlitz and LaBerge(1997) found the greatest effect for opposite-sex observer-subject pairs

Study One Results and Discussion

Participants

Thirty-two individuals took part as influencees (12 females 20 males) theirages ranged from 19 to 63 years (mean age 285 years)

Incomplete Data

One participant declined to give their age on the PIF form so the descriptivestatistics on age do not include this participant (so N = 31) Also due to a tech-nical hitch the exploratory variance scores were lost for one participant so N = 31 for this measure There were no drop-outs from this study

Remote Influence Measure

Table 1 presents a summary of the remote influence data in the form ofmean Percentage Score Index for co-operating and blocking conditions Therewas no significant difference between PSI for co-operating and blocking con-ditions (t = - 202 df = 31) Therefore Hypothesis 1 (predicting significantlygreater PSI scores for co-operating condition compared to blocking condition)was not supported by these results As planned the PSI scores for co-operatingand blocking runs were combined to examine Hypothesis 2 (of an overall re-mote mental influence effect in line with the calming and activating intentionsof the influencee) Mean PSI scores were exactly at chance expectation (t = - 031 df = 31) therefore there was no support for Hypothesis 2

Exploratory Questions

Variance Measure

Table 2 presents summary data for the exploratory measure of the amount ofvariance shown in each condition (lt 50 represents reduced variance gt 50 rep-resents increased variance) There was no significant difference between vari-ance scores for co-operating and blocking conditions (t = 835 df = 30) andwhen the conditions were combined variance did not depart from chance ex-pectation (t = - 244 df = 30) It is interesting to note however that the vari-ance measure gives a much wider range of scores than the PSI measure

Blocking vs Co-operating 523

TABLE 1Summary Data Mean PSI Scores

(N = 32 MCE = 50 range in brackets)

Co-operate PSI Block PSI Combined PSI

4996 5003 5000(4692- 5215) (4653- 5404) (4853- 5309)

TABLE 2Summary Data Mean Variance Scores(N = 31 MCE = 50 range in brackets)

Co-operate Variance Block Variance Combined Variance

5089 4864 4977(2648- 7012) (2492- 6484) (4116- 6297)

524 C Watt et al

suggesting that the variance measure may potentially be a more sensitive indi-cator of remote influence effect than PSI

Vividness of Mental Imagery

Low VVIQ scores represent participants reporting highly vivid imagery andhigh VVIQ scores represent those reporting little or no imagery The meanscore was 4275 (range from 21 to 80) Using a Pearson correlation we foundthat there was no evidence to suggest that those participants with low VVIQscores (high imagery) were better blockers than those with high scores (VVIQvs PSI blocking r = - 113) We also found that there was no evidence of a cor-relation between vividness of imagery and overall remote influence effectsusing the combined Percentage Score Index measure (VVIQ vs PSI overall r =- 077)

Influenceesrsquo Blocking and Co-operating Strategies

Influencees completed a questionnaire (after the session was over) askingthem to describe what strategies they used for blocking and co-operating Cat-egorizing these responses we found a difference in the kinds of strategies usedwhen blocking compared to co-operating as shown in Table 3 In the blockingcondition the majority of participants (469) used visual imagery strategiesFewer (281) participants used a combination of imagery and cognitivestrategies (such as repeating words related to blocking) while 25 reportedusing purely cognitive strategies for blocking (such as adopting a resistant anduncooperative attitude or thinking of irrelevancies) In the co-operating con-dition fewer participants used visual imagery alone (333) The majority ofparticipants (424) reported that their strategies for co-operating includedadopting a feeling of passiveness openness or relaxation whereas no partici-pants adopted this kind of strategy for the blocking condition 121 of partic-ipants used cognitive strategies alone (such as adopting a co-operative atti-tude) and the same proportion used a combination of visual imagery andcognitive strategies

Therefore there was a tendency for participants to use visual imagery forblocking and relaxationpassivity or visual imagery for co-operating This wasdespite the fact that we asked participants to attempt to adopt similar levels ofactivation for both co-operating and blocking strategies There was also a ten-dency not surprisingly for those participants with low VVIQ scores (ie vivid

TABLE 3Influenceesrsquo Mental Strategies ()

Vis Im Cognitive VisCog OpenRelax

Blocking 469 250 281 00Co-operating 333 121 121 424

imagers) to use visual imagery strategies Those who had high VVIQ scoresalso used visual imagery but more often used this in combination with othertactics

Other Qualitative Reports from Participants According to their post-ses-sion questionnaire responses participants generally felt comfortable about theattempt to be influenced (N = 24) Three participants said they felt apprehen-sive or uncomfortable with the idea of being influenced Interestingly of thesethree two were actively involved with parapsychology and were aware of theharmful psi implications discussed at the beginning of this report The thirdwas a psychology student who had attended a lecture given by Watt in whichthe possibilities of harmful psi were discussed Another group of three partici-pants said they would be uncomfortable with the idea of being influenced ifthey believed such influence was possible

Seven participants said they found it was difficult to sustain concentrationfor the 10 minutes that comprised each condition Five participants found thescreensaver distracting or disturbing Two participants specifically mentionedthat they found blocking a more taxing task than co-operating

Sex Effects

There were more males (N = 20) than females (N = 12) in this study howev-er there was no significant difference in PSI scores for remote physiology in-fluence between males and females (male PSI vs female PSI t = 1085 male PSI = 4987 female PSI = 5021)

Experimenter Effects

We studied whether there was any difference in PSI scores obtained by thethree experimenters The results are shown in Table 4 There seems to be nosign of experimenter effects in study one

InfluenceeInfluencer Sex Pairings

Table 5 gives the mean PSI scores for different influenceeinfluencer pair-ings Taking the combined PSI scores it can be seen that the highest remote in-fluence effect is for FM pairs followed by MF pairs followed by FF pairsand the lowest scoring is for MM pairs Post hoc it was found that the mixedsex pairings showed significantly higher remote influence scores than the same

Blocking vs Co-operating 525

TABLE 4Remote Influence Results by Experimenter

Experimenter N sessions Coop PSI Block PSI Combined PSI

CW 12 4983 5011 4997ZMcD 10 5015 4998 5006JR 10 4994 4999 4996

526 C Watt et al

sex pairings (t = - 2016 df = 30 p = 053 2 - t) This replicates the effect foundby Schlitz and LaBerge (1997)

Study One Conclusions

In Study 1 we developed a methodology to examine intentional resistance toa remote mental influence attempt (Watt et al 1997) We contrasted influ-enceesrsquo physiological activity when they were mentally co-operating with aremote attempt to influence their physiology with physiological activity wheninfluencees were mentally blocking the influence attempt We found no evi-dence for any remote influence effect overall and no difference betweenblocking and co-operating conditions Our qualitative measures yielded someinteresting information about participantsrsquo mental strategies and reactions dur-ing the remote influence session We found no difference between experi-menters in remote influence scores but we did find that mixed sex influ-enceeinfluencer pairings scored more highly than same sex pairings

Study Two Method

Study 2 was planned to explore in more depth the nature of the relationshipbetween aspects of personality and cognition and psi performance Partici-pants took part in three testing sessions approximately one week apart Thefirst consisted of various psychological measures the second consisted of aforced choice ESP task and the third consisted of a replication of the blocking-co-operating study (Study 1) described above The psychological measureswere not scored until after completion of the remote influence session Partici-pants received feedback about their forced choice ESP performance at the con-clusion of the second session For ease of interpretation the present paper isrestricted to a description of the third testing session which was essentially areplication of study one with fewer influencers and greater statistical power

We introduced some slight changes to the methodology of Study 2 based onour experiences with Study 1 At the start of each of the two runs in Study 1the participant was given auditory instructions either to ldquoblock nowrdquo or ldquoco-operate nowrdquo We observed that the auditory instructions appeared to affect theparticipantrsquos EDA as if they were startled As planned the first recordingepoch in each run was not used in analysis to give the participantrsquos EDA time

TABLE 5Mean PSI Scores for Different InfluenceeInfluencer Sex Pairings

Pairing N Coop PSI Block PSI Combined PSI

M-F 6 4932 5069 5001F-M 6 5043 5034 5048F-F 6 5022 4952 4993M-M 14 4988 4975 4981

to settle down However in order to remove any remaining potential artifactu-al effects on EDA in Study 2 only visual instructions were given

Another difference in methodology concerned the screensaver viewed bythe participants during the experimental session In Study 1 some participantsreported that the series of patterns produced by the computer were being dis-played at too fast a rate such that some participants found it distracting or dis-turbing Again we felt that this may have an artifactual effect on EDA so wedecided for study two to reduce the speed of the screensaver

Thirdly in Study 2 we did not give participants detailed information abouttheir remote influence performance until the feedback letters that were distrib-uted at the conclusion of the entire study The reason for this change was mere-ly to streamline the procedure on what was already a time-consuming experi-ment Finally in another streamlining change we did not take VVIQ measuresin Study 2

Planned Hypotheses and Analyses

These were the same as with Study 1 That is hypothesis 1 predicted higherPSI scores for the co-operating condition than for the blocking condition(measured with a related t-test one-tailed) If no significant difference wasfound between blocking PSI and co-operating PSI scores the two would becombined to look at overall remote influence effect (hypothesis 2 single-meant-test one-tailed) If a significant difference was found with hypothesis 1 theanalysis for hypothesis 2 would be conducted with the data for co-operatingsessions only

Exploratory Questions

As with study one we took an exploratory variance measure of remote EDAinfluence We also gathered qualitative descriptions of the mental strategiesparticipants adopted whilst blocking and co-operating We also looked at sexeffects experimenter effects and post hoc influenceeinfluencer sex pairings

Participants

Study 2 had two experimenterinfluencers (Watt amp Ravenscroft) After aparticipant was recruited (usually by Ravenscroft) they were assigned to oneof the two experimenters for all three testing sessions This assignment wasdone unsystematically according to the preferences and availability of the ex-perimenters As planned 50 volunteer participants were drawn from the gener-al participant pool at the Koestler Parapsychology Unit and by word ofmouth All participants completed the Participant Information Form and aConsent form prior to the study The participants consisted of 33 females and17 males Average age was 41 years range from 22 years to 76 years of age Inorder to increase the studyrsquos relevance to the general population undergradu-ate psychology students were not recruited as participants The main criterion

Blocking vs Co-operating 527

528 C Watt et al

for participation was that the participants were interested in and open-mindedabout parapsychology Again there were no drop-outs from this study Allparticipants received a detailed personal feedback letter at the conclusion ofthe study

Study Two Results and Discussion

Remote Influence Measure

PSI scores are summarized in Table 6 which shows that the mean PSI scorewhen co-operating was nonsignificantly lower than mean PSI score whenblocking (t = - 595 df = 49) Hypothesis 1 was therefore not supported Asplanned in the absence of any significant difference between blocking and co-operating PSI scores the two conditions were combined to test hypothesis 2which predicted overall significant remote mental influence performance Thecombined PSI score was significantly greater than mean chance expectation of50 (t = 1806 df = 49 p = 038 1 - t) Therefore hypothesis 2 was supportedgiving evidence of an overall remote influence effect upon psychophysiologyin the prespecified direction

The failure to support hypothesis 1 confirms what we found in Study 1 thatthere was no significant difference between blocking and co-operating condi-tions on remote influence scores Indeed both studies found a slight effect inthe opposite direction to that predicted that is slightly higher PSI scores whenblocking was the influenceersquos intention than when co-operating was their in-tention It was encouraging to find significant support for a remote influenceeffect in the predicted direction for blocking and co-operating conditions com-bined No such trend was found in the previous study though with a smallernumber of participants (N = 32) the earlier study had lower statistical powerthan the present study

Exploratory Questions

Variance Measure

Table 7 shows the results for the exploratory measure of EDA variance lt 50represents reduced variance gt 50 represents increased variance We found nodifference between variance scores for cooperate versus block conditions(t = 150 df = 49) When these are combined the resulting figure does not de-viate from chance expectation (t = 440 df = 49) Thus as in Study 1 it seems

TABLE 6Summary Data Mean PSI Scores

(N = 50 MCE = 50 range in brackets)

Co-operate PSI Block PSI Combined PSI

5008 5019 5016(4821- 5221) (4807- 5557) (4900- 5296)

that the exploratory variance measure does not reveal any additional indica-tion of the operation of a remote mental influence in this study However thewider range of variance scoring continues to suggest that this measure mayhave greater potential sensitivity than the summary PSI score

Influenceesrsquo Blocking and Co-operating Strategies

We were interested in learning more about what particular strategies partici-pants chose to adopt when blocking and co-operating As in Study 1 a post-session questionnaire was used to gather this information Table 8 shows thepercentage of participants who adopted each kind of strategy

The table shows a very similar pattern to that found in Study 1 The mainpoint of interest is that almost no participants took a passive approach toblocking (ie adopting an openrelaxing strategy) A much higher proportionpreferred to adopt this approach when co-operating The preferred mode forboth conditions was to use some form of visual imagery such as imagining ashield or a screen when blocking and imagining an open channel such as ariver when co-operating

The majority of participants reported no concerns over the possibility ofhaving their physiology remotely influenced by another person as found instudy one

Sex Effects

As in study one the sexes were not equally represented among participantsWhereas Study 1 had an excess of male participants in Study 2 twice as manyfemales (N = 33) as males (N = 17) participated It was therefore important tolook for sex differences since any such differences could confound the inter-pretation of overall trends We found no difference between sexes when co-op-erating (female mean PSI score = 5008 male mean PSI score = 5006 t = 107 df = 48) There was a nonsignificant trend for females to have higher

Blocking vs Co-operating 529

TABLE 7Summary Data Mean Variance Scores(N = 50 MCE = 50 range in brackets)

Co-operate Variance Block Variance Combined Variance

5076 5054 5034(3164- 7279) (3572- 6813) (3749- 6671)

TABLE 8Influenceesrsquo Mental Strategies ()

Visual Imagery Cognitive VisCog OpenRelax

Blocking 64 26 6 2Co-operating 60 8 12 20

530 C Watt et al

PSI scores than males when blocking (mean PSI scores 5030 and 4998 re-spectively t = 992 df = 48) Combining blocking and co-operating conditionsthere was no overall difference between the sexes (mean PSI scores 5020 forfemales 5008 for males t = 665 df = 48)

Experimenter Effects

Study 2 found a statistically significant overall remote influence effect(combined PSI = 50155 t = 1806) Table 9 shows the breakdown of remoteinfluence scores for the two experimenters CW and JR Looking at the overalldata for the two experimenters we see that CWrsquos sessions were associatedwith most of the remote influence effect However due to the smaller N CWrsquossessions are not independently significant (t = 1632) and there is not a signifi-cant difference between CW and JRrsquos overall PSI scores (t = 678) Post hocwe calculated effect size r according to the formula r = [t2(t2+df)]12 for the twoexperimenters For JR (t = 817 df = 22) we found a small effect sizer = +17 For CW (t = 1632 df = 26) the effect size was moderate r = +30This may be some post hoc indication of an experimenter effect in Study 2

InfluenceeInfluencer Sex Pairings

As can be seen from Table 10 Study 2 found exactly the opposite trend tothat found in Study 1 On a post hoc test same sex pairings scored significant-ly higher on the remote influence PSI measure than mixed sex pairings(t = - 2034 df = 48 p = 048 2 - t) This contradicts Schlitz and LaBergersquos(1997) finding of greater autonomic detection of remote staring for mixed sexpairs

TABLE 9Remote Influence Results by Experimenter

Experimenter N sessions Coop PSI Block PSI Combined PSI

CW 27 5011 5026 5021JR 23 5003 5010 5009

TABLE 10Mean PSI Scores for Different InfluenceeInfluencer Sex Pairings

Pairing N Coop PSI Block PSI Combined PSI

MF 11 4977 5009 4997FM 17 4983 5021 5003FF 16 5035 5038 5038MM 6 5060 4978 5027

General Summary and Conclusions

We have described the findings of two studies exploring the limits of remotemental influence effects We wished to see whether the mental strategiesadopted by the influencee would affect remote influence scores In particularcould an influencee block or resist a remote influence attempt We recorded in-fluenceesrsquo electrodermal activity while the influencer attempted to calm to ac-tivate or to have no influence over the influenceesrsquo physiology In each studythere were two conditions reflecting the influenceesrsquo designated mental strate-gies in response to the remote influence attempt blocking and co-operating

There were only two formal hypotheses 1) that remote influence scoreswould be significantly higher during the co-operating condition than duringthe blocking condition and 2) that overall there would be significant evidenceof remote mental influence on physiology in the designated direction (iecalming when calming was the influencerrsquos intention activating when activat-ing was the influencerrsquos intention) Contrary to prediction Study 1 found thatremote influence effects were higher in the block condition than in the co-op-erate condition (but not to a significant extent) Also overall remote influenceperformance was at chance level This study therefore found no support for ei-ther of the formal hypotheses Study 2 consisted of a replication of Study 1with a few minor adjustments plus additional measures that were not reportedin this paper Again contrary to prediction remote influence scores were non-significantly higher in the block condition than in the co-operate conditionThis replicates the trend found in Study 1 failing to support hypothesis oneThese results suggest that the mental strategy adopted by the influencee in theface of a remote influence attempt has no effect on remote influence perfor-mance Referring back to the questions posed in the introduction then wehave not found any evidence to suggest that it is possible to resist a remote in-fluence effect and there is as yet no indication of the possible limiting condi-tions for remote mental influence

Unlike Study 1 Study 2 found statistically significant evidence of an overallremote influence effect in the prespecified direction (ie calming of EDAwhen that was the influencerrsquos intention activation of EDA when that was theinfluencerrsquos intention) In order to facilitate comparison with a recent meta-an-alytic review of remote influence studies we calculated the effect size (r) ac-cording to the formula r = [t2(t2+df)]12 For their review of 19 EDA influencestudies Schlitz and Braud (1997) found an average effect size of r = +25 OurStudy 1 with overall t = - 031 df = 31 gives a negligible effect size of r = - 006 Study 2 with overall t = 1806 df = 49 gives r = +25 Study 2 there-fore replicates the mean effect size found in EDA studies conducted by otherresearchers in other parapsychology laboratories Post hoc we can combinethe data of the two studies for greater statistical power Overall t = 1176 df = 81 giving a small effect size of r = +13

We can find no clear reason why Study 2 found an EDA influence effectwhile Study 1 did not A number of factors are worth considering 1) The

Blocking vs Co-operating 531

532 C Watt et al

significant result could be due to chance or undetected artifact 2) Study 2 im-plemented what we considered to be minor methodological improvements thatmight have removed some artifactual fluctuations from the EDA data that weobserved in study one 3) Each experimenter in Study 2 conducted three testingsessions with their participant By the third session in which the remote influ-ence measure was taken experimenters were on quite friendly terms with theirparticipants perhaps more so than in Study 1 where each participant came infor only one testing session 4) The participants in Study 2 were more represen-tative of the general population we felt than those in Study 1 Study 1 in-volved more undergraduate psychology students and associates of the experi-menters 5) The experimenters in study one were less experienced with theremote influence protocol than when they conducted Study 2

In addition to the formal hypotheses we reported a number of exploratorymeasures Study 1 found that influenceesrsquo self-reported vividness of visual im-agery appeared to be unrelated to the remote influence performance In bothstudies influenceesrsquo qualitative descriptions of the actual mental strategiesthey employed when blocking and co-operating revealed a tendency for differ-ent strategies to be preferred for the different conditions The use of visual im-agery was popular for both blocking (eg visualizing being surrounded by aprotective cocoon) and co-operating (eg visualizing a river flowing betweeninfluencer and influencee) However a considerable proportion of participantsreported adopting an open or relaxed frame of mind when co-operating whilstalmost no participants chose this strategy when blocking This was despite thefact that prior to each session we encouraged participants to adopt comparablestrategies for blocking and co-operating

It was interesting and perhaps surprising to note that in both studies themajority of participants reported no feelings of concern or discomfort at thepossibility of having their physiology remotely influenced Those who wereconcerned tended to be those who knew more about parapsychology and thepossible harmful applications of remote influence A small proportion of par-ticipants was unconcerned because they did not believe such an influence waspossible in the first place

Both studies employed an exploratory variance measure of EDA influenceThis measure showed a greater range of scoring overall thus perhaps greaterpotential sensitivity to a remote influence effect However the variance mea-sure showed no indication of a remote influence effect To date researchershave employed a variety of indicators of remote influence as the literaturegrows progress may be made on identifying which is the most appropriate in-dicator

Neither study found significant sex differences in remote influence effectsthough in both studies female influencees tended to have higher remote influ-ence scores than male influencees

In terms of their overall remote influence effects there appeared to be nodifference between experimenterinfluencers in Study 1 (r = +06 - 05 and

- 06) In Study 2 effect size measures indicated that the experimenters dif-fered in remote influence effect sizes (r = +17 and +30) Due to the smaller Nthis difference was not found to be statistically significant Post hoc combin-ing Studies 1 and 2 for experimenters JW and CW (ZMcD was not an experi-menter in Study 2) the effect sizes are r = +08 and r = +20 respectively

We conducted a post hoc analysis of the effects of influenceeinfluencer sexpairings on remote influence scores Our two studies came up with contradic-tory findings each statistically significant Study one found that mixed sexpairings had significantly higher remote influence scores than same sex pair-ings This replicated the post hoc finding of Schlitz and LaBerge (1997) in con-nection with the autonomic detection of remote observation Study 2 foundthat same sex pairings had significantly higher remote influence scores thanmixed sex pairings Our contradictory findings do not help to clarify the pic-ture on the importance or otherwise of sex pairings It is quite possible that sexis not the most salient aspect of the dynamic between pairs Age personalityand familiarity are likely to be important variables moderating the relationshipbetween influencer and influencee

There are many more variables that we could look at but as the number ofanalyses multiplies so does the opportunity for spurious significant resultsFor this reason we have named only two formal hypotheses for each experi-ment All other analyses are regarded as exploratory and any trends that wehave highlighted above would require replication

In conclusion these two studies have found no evidence of a difference inremote influence effects when influencees are asked to adopt cooperative men-tal strategies compared to when they are asked to adopt blocking mental strate-gies In fact contrary to expectation both studies found slightly higher remoteinfluence effects for the blocking condition compared to the co-operating con-dition On this basis we have found no indication of limiting conditions for re-mote influence effects Therefore while anecdotal advice exists on how to de-fend oneself against a remote physiological influence attempt in the realworld laboratory evidence to date has provided no support for this advice

Given the importance of this question both in practical terms and in theoret-ical terms we argue that this line of research deserves continued investigationIf a way could be found to conduct research with a more realistic scenario mdash where the remote influence was not just neutrally and abstractlydefined as unwelcome mdash then our research might have more meaningful thingsto say about the parameters of remote influence It is interesting and perhapssurprising to note that when Watt has lectured on this topic to informed lay au-diences the feedback has often been that researchers must use more potential-ly harmful or disruptive remote influence (of course with the informed consentof the participants) While there are clear ethical inhibitions when consideringpossible physiological influence perhaps research using higher level influencewould be more acceptable For example one could attempt to disrupt or en-hance the concentration of chess players who were aware of the influence

Blocking vs Co-operating 533

534 C Watt et al

attempt and who were concurrently blocking or co-operating with the influ-ence attempt Research already exists to suggest that remote influence does notjust operate at physiological levels Attention focusing for example can be fa-cilitated by a remote helper (Brady amp Morris 1997 Braud et al 1995) Futureresearch in this area must therefore expand beyond physiological influences

Acknowledgements

The authors gratefully acknowledge the financial support they receivedfrom the Institut fuumlr Grenzgebiete der Psychologie und PsychohygieneFreiburg Germany while conducting these studies We would also like tothank Dr Paul Stevens for generously helping to solve software difficultieswhen preparing for these studies and the anonymous reviewer for helpfulcomments Address reprint requests to Dr C Watt Department of PsychologyUniversity of Edinburgh 7 George Square Edinburgh EH8 9JZ Scotland UK

ReferencesBrady C amp Morris R L (1997) Attention focusing facilitated through remote mental interac-

tion A replication and exploration of parameters In R Wiseman (Ed) Proceedings of Pre-sented Papers The Parapsychological Association 40th Annual Convention held in conjunc-tion with The Society for Psychical Research pp 73ndash91

Braud W G (1985) Blockingshielding psychic functioning through psychological and psychictechniques A report of three preliminary studies In RA White and J Solfvin (Eds) Re-search in Parapsychology 1984 Metuchen NJ Scarecrow Press pp 42ndash44

Braud W G amp Schlitz M J (1991) Consciousness interactions with remote biological systemsAnomalous intentionality effects Subtle Energies 2 1

Braud W G Schlitz M J Collins J and Klitch H (1985) Further studies of the bio-PK effectFeedback blocking specificitygenerality In RA White amp J Solfvin (Eds) Research inParapsychology 1984 Metuchen NJ Scarecrow Press pp 45ndash48

Braud W G Shafer D McNeill K and Guerra V (1995) Attention focusing facilitatedthrough remote mental interaction Journal of the American Society for Psychical Research89 103

Cohen J (1997) Statistical Power Analysis for the Behavioral Sciences New York AcademicPress

Dalton K (1994) Remotely influenced ESP performance in a computer task A preliminarystudy Proceedings of the 37th Annual Convention of the Parapsychological Association pp95ndash103

Dossey L (1994) Healing and the mind Is there a dark side Journal of Scientific Exploration 873

Halifax-Grof J (1974) Hex death In A Angoff amp D Barth (Eds) Parapsychology and Anthro-pology Proceedings of an International Conference Held in London England August 29ndash311973 NY Parapsychology Foundation

Hope M (1983) Practical Techniques of Psychic Self-Defence Wellingborough Northampton-shire Aquarian Press

Marks DF (1973) Visual imagery differences in the recall of pictures British Journal of Psy-chology 64 17

Nash CB (1982) Psychokinetic control of bacterial growth Journal of the Society for PsychicalResearch 51 217

Radin DI Taylor RK and Braud WG (1995) Remote mental influence of human electroder-mal activity A pilot replication European Journal of Parapsychology 11 19

Schlitz M amp Braud W G (1997) Distant intentionality and healing Assessing the evidence Al-ternative Therapies 3 62

Schlitz M J amp LaBerge S (1997) Covert observation increases skin conductance in subjectsunaware of when they are being observed A replication Journal of Parapsychology 61 185

Schmeidler G R (1958) Agent-percipient relationships Journal of the American Society forPsychical Research 52 47

Schmeidler G R (1961) Evidence for two kinds of telepathy International Journal of Parapsy-chology 3 5

Tart C T (1979) A survey of expert opinion on potentially negative uses of psi United Statesgovernment interest in psi and the level of research funding of the field In WG Roll (Ed) Re-search in Parapsychology 1978 Metuchen NJ Scarecrow Press pp 54

Tart C T amp Labore C M (1986) Attitudes towards strongly functioning psi A preliminary sur-vey Journal of the America Society for Psychical Research 80 163

Tart C T (1986) Psychicsrsquo fears of psychic powers Journal of the American Society for Psychi-cal Research 80 278

Watt C A McDermott Z amp Ravenscroft J (1997) Exploring the limits of remote mental inter-action A comparison of blocking versus co-operating strategies In R Wiseman (Ed) Pro-ceedings of the 40th Annual Convention of the Parapsychological Association held in conjunc-tion with the Society for Psychical Research Brighton England pp 464ndash476

Blocking vs Co-operating 535

522 C Watt et al

to the PSI score we looked at the amount of variance that was shown in block-ing and co-operating scores The variance measure was calculated as followsMean values for activate periods and calm periods were calculated separatelyfor block and co-operate conditions Mean values were taken from each indi-vidual score in that condition and squared and summed S [mean value for con-dition - individual score]2 This gave a summation of variability around themean for each activate calm and rest period for block and cooperate condi-tions It was used in conjunction with the mean level of EDA activity givingan indication of the motion of the wave (ie the variability of the EDA activi-ty) rather than the overall level

VVIQ

Anecdotally the use of appropriate imagery (eg visualize being surroundedby a shielding white halo of white light) is suggested to be an effective blockingtechnique (eg Hope 1983) We therefore wished to see if participants who re-ported experiencing vivid imagery with ease are perhaps more effective block-ers than those who do not report clear mental imagery Descriptive data wasgathered to explore any trends in PSI scores for blockingco-operating forthose participants with VVIQ scores ranging from low (ie reporting vividmental imagery) to high (ie reporting little or no mental imagery)

Qualitative Reports

We also wanted to report data illustrating what kind of different strategiesparticipants used when they were blocking and co-operating These qualitativereports were gathered in the form of the participantsrsquo written responses toopen-ended questions on a post-session questionnaire Watt and Ravenscroftindependently coded these written responses into categories and cross-checked to ensure coding agreement

Sex and Experimenter Effects

We also examined the data for sex effects and experimenter effects becausethese variables could confound the interpretation of any overall effects Final-ly a post hoc analysis was conducted on influenceeinfluencer sex pairings Ina study on the unconscious detection of remote staring Schlitz and LaBerge(1997) found the greatest effect for opposite-sex observer-subject pairs

Study One Results and Discussion

Participants

Thirty-two individuals took part as influencees (12 females 20 males) theirages ranged from 19 to 63 years (mean age 285 years)

Incomplete Data

One participant declined to give their age on the PIF form so the descriptivestatistics on age do not include this participant (so N = 31) Also due to a tech-nical hitch the exploratory variance scores were lost for one participant so N = 31 for this measure There were no drop-outs from this study

Remote Influence Measure

Table 1 presents a summary of the remote influence data in the form ofmean Percentage Score Index for co-operating and blocking conditions Therewas no significant difference between PSI for co-operating and blocking con-ditions (t = - 202 df = 31) Therefore Hypothesis 1 (predicting significantlygreater PSI scores for co-operating condition compared to blocking condition)was not supported by these results As planned the PSI scores for co-operatingand blocking runs were combined to examine Hypothesis 2 (of an overall re-mote mental influence effect in line with the calming and activating intentionsof the influencee) Mean PSI scores were exactly at chance expectation (t = - 031 df = 31) therefore there was no support for Hypothesis 2

Exploratory Questions

Variance Measure

Table 2 presents summary data for the exploratory measure of the amount ofvariance shown in each condition (lt 50 represents reduced variance gt 50 rep-resents increased variance) There was no significant difference between vari-ance scores for co-operating and blocking conditions (t = 835 df = 30) andwhen the conditions were combined variance did not depart from chance ex-pectation (t = - 244 df = 30) It is interesting to note however that the vari-ance measure gives a much wider range of scores than the PSI measure

Blocking vs Co-operating 523

TABLE 1Summary Data Mean PSI Scores

(N = 32 MCE = 50 range in brackets)

Co-operate PSI Block PSI Combined PSI

4996 5003 5000(4692- 5215) (4653- 5404) (4853- 5309)

TABLE 2Summary Data Mean Variance Scores(N = 31 MCE = 50 range in brackets)

Co-operate Variance Block Variance Combined Variance

5089 4864 4977(2648- 7012) (2492- 6484) (4116- 6297)

524 C Watt et al

suggesting that the variance measure may potentially be a more sensitive indi-cator of remote influence effect than PSI

Vividness of Mental Imagery

Low VVIQ scores represent participants reporting highly vivid imagery andhigh VVIQ scores represent those reporting little or no imagery The meanscore was 4275 (range from 21 to 80) Using a Pearson correlation we foundthat there was no evidence to suggest that those participants with low VVIQscores (high imagery) were better blockers than those with high scores (VVIQvs PSI blocking r = - 113) We also found that there was no evidence of a cor-relation between vividness of imagery and overall remote influence effectsusing the combined Percentage Score Index measure (VVIQ vs PSI overall r =- 077)

Influenceesrsquo Blocking and Co-operating Strategies

Influencees completed a questionnaire (after the session was over) askingthem to describe what strategies they used for blocking and co-operating Cat-egorizing these responses we found a difference in the kinds of strategies usedwhen blocking compared to co-operating as shown in Table 3 In the blockingcondition the majority of participants (469) used visual imagery strategiesFewer (281) participants used a combination of imagery and cognitivestrategies (such as repeating words related to blocking) while 25 reportedusing purely cognitive strategies for blocking (such as adopting a resistant anduncooperative attitude or thinking of irrelevancies) In the co-operating con-dition fewer participants used visual imagery alone (333) The majority ofparticipants (424) reported that their strategies for co-operating includedadopting a feeling of passiveness openness or relaxation whereas no partici-pants adopted this kind of strategy for the blocking condition 121 of partic-ipants used cognitive strategies alone (such as adopting a co-operative atti-tude) and the same proportion used a combination of visual imagery andcognitive strategies

Therefore there was a tendency for participants to use visual imagery forblocking and relaxationpassivity or visual imagery for co-operating This wasdespite the fact that we asked participants to attempt to adopt similar levels ofactivation for both co-operating and blocking strategies There was also a ten-dency not surprisingly for those participants with low VVIQ scores (ie vivid

TABLE 3Influenceesrsquo Mental Strategies ()

Vis Im Cognitive VisCog OpenRelax

Blocking 469 250 281 00Co-operating 333 121 121 424

imagers) to use visual imagery strategies Those who had high VVIQ scoresalso used visual imagery but more often used this in combination with othertactics

Other Qualitative Reports from Participants According to their post-ses-sion questionnaire responses participants generally felt comfortable about theattempt to be influenced (N = 24) Three participants said they felt apprehen-sive or uncomfortable with the idea of being influenced Interestingly of thesethree two were actively involved with parapsychology and were aware of theharmful psi implications discussed at the beginning of this report The thirdwas a psychology student who had attended a lecture given by Watt in whichthe possibilities of harmful psi were discussed Another group of three partici-pants said they would be uncomfortable with the idea of being influenced ifthey believed such influence was possible

Seven participants said they found it was difficult to sustain concentrationfor the 10 minutes that comprised each condition Five participants found thescreensaver distracting or disturbing Two participants specifically mentionedthat they found blocking a more taxing task than co-operating

Sex Effects

There were more males (N = 20) than females (N = 12) in this study howev-er there was no significant difference in PSI scores for remote physiology in-fluence between males and females (male PSI vs female PSI t = 1085 male PSI = 4987 female PSI = 5021)

Experimenter Effects

We studied whether there was any difference in PSI scores obtained by thethree experimenters The results are shown in Table 4 There seems to be nosign of experimenter effects in study one

InfluenceeInfluencer Sex Pairings

Table 5 gives the mean PSI scores for different influenceeinfluencer pair-ings Taking the combined PSI scores it can be seen that the highest remote in-fluence effect is for FM pairs followed by MF pairs followed by FF pairsand the lowest scoring is for MM pairs Post hoc it was found that the mixedsex pairings showed significantly higher remote influence scores than the same

Blocking vs Co-operating 525

TABLE 4Remote Influence Results by Experimenter

Experimenter N sessions Coop PSI Block PSI Combined PSI

CW 12 4983 5011 4997ZMcD 10 5015 4998 5006JR 10 4994 4999 4996

526 C Watt et al

sex pairings (t = - 2016 df = 30 p = 053 2 - t) This replicates the effect foundby Schlitz and LaBerge (1997)

Study One Conclusions

In Study 1 we developed a methodology to examine intentional resistance toa remote mental influence attempt (Watt et al 1997) We contrasted influ-enceesrsquo physiological activity when they were mentally co-operating with aremote attempt to influence their physiology with physiological activity wheninfluencees were mentally blocking the influence attempt We found no evi-dence for any remote influence effect overall and no difference betweenblocking and co-operating conditions Our qualitative measures yielded someinteresting information about participantsrsquo mental strategies and reactions dur-ing the remote influence session We found no difference between experi-menters in remote influence scores but we did find that mixed sex influ-enceeinfluencer pairings scored more highly than same sex pairings

Study Two Method

Study 2 was planned to explore in more depth the nature of the relationshipbetween aspects of personality and cognition and psi performance Partici-pants took part in three testing sessions approximately one week apart Thefirst consisted of various psychological measures the second consisted of aforced choice ESP task and the third consisted of a replication of the blocking-co-operating study (Study 1) described above The psychological measureswere not scored until after completion of the remote influence session Partici-pants received feedback about their forced choice ESP performance at the con-clusion of the second session For ease of interpretation the present paper isrestricted to a description of the third testing session which was essentially areplication of study one with fewer influencers and greater statistical power

We introduced some slight changes to the methodology of Study 2 based onour experiences with Study 1 At the start of each of the two runs in Study 1the participant was given auditory instructions either to ldquoblock nowrdquo or ldquoco-operate nowrdquo We observed that the auditory instructions appeared to affect theparticipantrsquos EDA as if they were startled As planned the first recordingepoch in each run was not used in analysis to give the participantrsquos EDA time

TABLE 5Mean PSI Scores for Different InfluenceeInfluencer Sex Pairings

Pairing N Coop PSI Block PSI Combined PSI

M-F 6 4932 5069 5001F-M 6 5043 5034 5048F-F 6 5022 4952 4993M-M 14 4988 4975 4981

to settle down However in order to remove any remaining potential artifactu-al effects on EDA in Study 2 only visual instructions were given

Another difference in methodology concerned the screensaver viewed bythe participants during the experimental session In Study 1 some participantsreported that the series of patterns produced by the computer were being dis-played at too fast a rate such that some participants found it distracting or dis-turbing Again we felt that this may have an artifactual effect on EDA so wedecided for study two to reduce the speed of the screensaver

Thirdly in Study 2 we did not give participants detailed information abouttheir remote influence performance until the feedback letters that were distrib-uted at the conclusion of the entire study The reason for this change was mere-ly to streamline the procedure on what was already a time-consuming experi-ment Finally in another streamlining change we did not take VVIQ measuresin Study 2

Planned Hypotheses and Analyses

These were the same as with Study 1 That is hypothesis 1 predicted higherPSI scores for the co-operating condition than for the blocking condition(measured with a related t-test one-tailed) If no significant difference wasfound between blocking PSI and co-operating PSI scores the two would becombined to look at overall remote influence effect (hypothesis 2 single-meant-test one-tailed) If a significant difference was found with hypothesis 1 theanalysis for hypothesis 2 would be conducted with the data for co-operatingsessions only

Exploratory Questions

As with study one we took an exploratory variance measure of remote EDAinfluence We also gathered qualitative descriptions of the mental strategiesparticipants adopted whilst blocking and co-operating We also looked at sexeffects experimenter effects and post hoc influenceeinfluencer sex pairings

Participants

Study 2 had two experimenterinfluencers (Watt amp Ravenscroft) After aparticipant was recruited (usually by Ravenscroft) they were assigned to oneof the two experimenters for all three testing sessions This assignment wasdone unsystematically according to the preferences and availability of the ex-perimenters As planned 50 volunteer participants were drawn from the gener-al participant pool at the Koestler Parapsychology Unit and by word ofmouth All participants completed the Participant Information Form and aConsent form prior to the study The participants consisted of 33 females and17 males Average age was 41 years range from 22 years to 76 years of age Inorder to increase the studyrsquos relevance to the general population undergradu-ate psychology students were not recruited as participants The main criterion

Blocking vs Co-operating 527

528 C Watt et al

for participation was that the participants were interested in and open-mindedabout parapsychology Again there were no drop-outs from this study Allparticipants received a detailed personal feedback letter at the conclusion ofthe study

Study Two Results and Discussion

Remote Influence Measure

PSI scores are summarized in Table 6 which shows that the mean PSI scorewhen co-operating was nonsignificantly lower than mean PSI score whenblocking (t = - 595 df = 49) Hypothesis 1 was therefore not supported Asplanned in the absence of any significant difference between blocking and co-operating PSI scores the two conditions were combined to test hypothesis 2which predicted overall significant remote mental influence performance Thecombined PSI score was significantly greater than mean chance expectation of50 (t = 1806 df = 49 p = 038 1 - t) Therefore hypothesis 2 was supportedgiving evidence of an overall remote influence effect upon psychophysiologyin the prespecified direction

The failure to support hypothesis 1 confirms what we found in Study 1 thatthere was no significant difference between blocking and co-operating condi-tions on remote influence scores Indeed both studies found a slight effect inthe opposite direction to that predicted that is slightly higher PSI scores whenblocking was the influenceersquos intention than when co-operating was their in-tention It was encouraging to find significant support for a remote influenceeffect in the predicted direction for blocking and co-operating conditions com-bined No such trend was found in the previous study though with a smallernumber of participants (N = 32) the earlier study had lower statistical powerthan the present study

Exploratory Questions

Variance Measure

Table 7 shows the results for the exploratory measure of EDA variance lt 50represents reduced variance gt 50 represents increased variance We found nodifference between variance scores for cooperate versus block conditions(t = 150 df = 49) When these are combined the resulting figure does not de-viate from chance expectation (t = 440 df = 49) Thus as in Study 1 it seems

TABLE 6Summary Data Mean PSI Scores

(N = 50 MCE = 50 range in brackets)

Co-operate PSI Block PSI Combined PSI

5008 5019 5016(4821- 5221) (4807- 5557) (4900- 5296)

that the exploratory variance measure does not reveal any additional indica-tion of the operation of a remote mental influence in this study However thewider range of variance scoring continues to suggest that this measure mayhave greater potential sensitivity than the summary PSI score

Influenceesrsquo Blocking and Co-operating Strategies

We were interested in learning more about what particular strategies partici-pants chose to adopt when blocking and co-operating As in Study 1 a post-session questionnaire was used to gather this information Table 8 shows thepercentage of participants who adopted each kind of strategy

The table shows a very similar pattern to that found in Study 1 The mainpoint of interest is that almost no participants took a passive approach toblocking (ie adopting an openrelaxing strategy) A much higher proportionpreferred to adopt this approach when co-operating The preferred mode forboth conditions was to use some form of visual imagery such as imagining ashield or a screen when blocking and imagining an open channel such as ariver when co-operating

The majority of participants reported no concerns over the possibility ofhaving their physiology remotely influenced by another person as found instudy one

Sex Effects

As in study one the sexes were not equally represented among participantsWhereas Study 1 had an excess of male participants in Study 2 twice as manyfemales (N = 33) as males (N = 17) participated It was therefore important tolook for sex differences since any such differences could confound the inter-pretation of overall trends We found no difference between sexes when co-op-erating (female mean PSI score = 5008 male mean PSI score = 5006 t = 107 df = 48) There was a nonsignificant trend for females to have higher

Blocking vs Co-operating 529

TABLE 7Summary Data Mean Variance Scores(N = 50 MCE = 50 range in brackets)

Co-operate Variance Block Variance Combined Variance

5076 5054 5034(3164- 7279) (3572- 6813) (3749- 6671)

TABLE 8Influenceesrsquo Mental Strategies ()

Visual Imagery Cognitive VisCog OpenRelax

Blocking 64 26 6 2Co-operating 60 8 12 20

530 C Watt et al

PSI scores than males when blocking (mean PSI scores 5030 and 4998 re-spectively t = 992 df = 48) Combining blocking and co-operating conditionsthere was no overall difference between the sexes (mean PSI scores 5020 forfemales 5008 for males t = 665 df = 48)

Experimenter Effects

Study 2 found a statistically significant overall remote influence effect(combined PSI = 50155 t = 1806) Table 9 shows the breakdown of remoteinfluence scores for the two experimenters CW and JR Looking at the overalldata for the two experimenters we see that CWrsquos sessions were associatedwith most of the remote influence effect However due to the smaller N CWrsquossessions are not independently significant (t = 1632) and there is not a signifi-cant difference between CW and JRrsquos overall PSI scores (t = 678) Post hocwe calculated effect size r according to the formula r = [t2(t2+df)]12 for the twoexperimenters For JR (t = 817 df = 22) we found a small effect sizer = +17 For CW (t = 1632 df = 26) the effect size was moderate r = +30This may be some post hoc indication of an experimenter effect in Study 2

InfluenceeInfluencer Sex Pairings

As can be seen from Table 10 Study 2 found exactly the opposite trend tothat found in Study 1 On a post hoc test same sex pairings scored significant-ly higher on the remote influence PSI measure than mixed sex pairings(t = - 2034 df = 48 p = 048 2 - t) This contradicts Schlitz and LaBergersquos(1997) finding of greater autonomic detection of remote staring for mixed sexpairs

TABLE 9Remote Influence Results by Experimenter

Experimenter N sessions Coop PSI Block PSI Combined PSI

CW 27 5011 5026 5021JR 23 5003 5010 5009

TABLE 10Mean PSI Scores for Different InfluenceeInfluencer Sex Pairings

Pairing N Coop PSI Block PSI Combined PSI

MF 11 4977 5009 4997FM 17 4983 5021 5003FF 16 5035 5038 5038MM 6 5060 4978 5027

General Summary and Conclusions

We have described the findings of two studies exploring the limits of remotemental influence effects We wished to see whether the mental strategiesadopted by the influencee would affect remote influence scores In particularcould an influencee block or resist a remote influence attempt We recorded in-fluenceesrsquo electrodermal activity while the influencer attempted to calm to ac-tivate or to have no influence over the influenceesrsquo physiology In each studythere were two conditions reflecting the influenceesrsquo designated mental strate-gies in response to the remote influence attempt blocking and co-operating

There were only two formal hypotheses 1) that remote influence scoreswould be significantly higher during the co-operating condition than duringthe blocking condition and 2) that overall there would be significant evidenceof remote mental influence on physiology in the designated direction (iecalming when calming was the influencerrsquos intention activating when activat-ing was the influencerrsquos intention) Contrary to prediction Study 1 found thatremote influence effects were higher in the block condition than in the co-op-erate condition (but not to a significant extent) Also overall remote influenceperformance was at chance level This study therefore found no support for ei-ther of the formal hypotheses Study 2 consisted of a replication of Study 1with a few minor adjustments plus additional measures that were not reportedin this paper Again contrary to prediction remote influence scores were non-significantly higher in the block condition than in the co-operate conditionThis replicates the trend found in Study 1 failing to support hypothesis oneThese results suggest that the mental strategy adopted by the influencee in theface of a remote influence attempt has no effect on remote influence perfor-mance Referring back to the questions posed in the introduction then wehave not found any evidence to suggest that it is possible to resist a remote in-fluence effect and there is as yet no indication of the possible limiting condi-tions for remote mental influence

Unlike Study 1 Study 2 found statistically significant evidence of an overallremote influence effect in the prespecified direction (ie calming of EDAwhen that was the influencerrsquos intention activation of EDA when that was theinfluencerrsquos intention) In order to facilitate comparison with a recent meta-an-alytic review of remote influence studies we calculated the effect size (r) ac-cording to the formula r = [t2(t2+df)]12 For their review of 19 EDA influencestudies Schlitz and Braud (1997) found an average effect size of r = +25 OurStudy 1 with overall t = - 031 df = 31 gives a negligible effect size of r = - 006 Study 2 with overall t = 1806 df = 49 gives r = +25 Study 2 there-fore replicates the mean effect size found in EDA studies conducted by otherresearchers in other parapsychology laboratories Post hoc we can combinethe data of the two studies for greater statistical power Overall t = 1176 df = 81 giving a small effect size of r = +13

We can find no clear reason why Study 2 found an EDA influence effectwhile Study 1 did not A number of factors are worth considering 1) The

Blocking vs Co-operating 531

532 C Watt et al

significant result could be due to chance or undetected artifact 2) Study 2 im-plemented what we considered to be minor methodological improvements thatmight have removed some artifactual fluctuations from the EDA data that weobserved in study one 3) Each experimenter in Study 2 conducted three testingsessions with their participant By the third session in which the remote influ-ence measure was taken experimenters were on quite friendly terms with theirparticipants perhaps more so than in Study 1 where each participant came infor only one testing session 4) The participants in Study 2 were more represen-tative of the general population we felt than those in Study 1 Study 1 in-volved more undergraduate psychology students and associates of the experi-menters 5) The experimenters in study one were less experienced with theremote influence protocol than when they conducted Study 2

In addition to the formal hypotheses we reported a number of exploratorymeasures Study 1 found that influenceesrsquo self-reported vividness of visual im-agery appeared to be unrelated to the remote influence performance In bothstudies influenceesrsquo qualitative descriptions of the actual mental strategiesthey employed when blocking and co-operating revealed a tendency for differ-ent strategies to be preferred for the different conditions The use of visual im-agery was popular for both blocking (eg visualizing being surrounded by aprotective cocoon) and co-operating (eg visualizing a river flowing betweeninfluencer and influencee) However a considerable proportion of participantsreported adopting an open or relaxed frame of mind when co-operating whilstalmost no participants chose this strategy when blocking This was despite thefact that prior to each session we encouraged participants to adopt comparablestrategies for blocking and co-operating

It was interesting and perhaps surprising to note that in both studies themajority of participants reported no feelings of concern or discomfort at thepossibility of having their physiology remotely influenced Those who wereconcerned tended to be those who knew more about parapsychology and thepossible harmful applications of remote influence A small proportion of par-ticipants was unconcerned because they did not believe such an influence waspossible in the first place

Both studies employed an exploratory variance measure of EDA influenceThis measure showed a greater range of scoring overall thus perhaps greaterpotential sensitivity to a remote influence effect However the variance mea-sure showed no indication of a remote influence effect To date researchershave employed a variety of indicators of remote influence as the literaturegrows progress may be made on identifying which is the most appropriate in-dicator

Neither study found significant sex differences in remote influence effectsthough in both studies female influencees tended to have higher remote influ-ence scores than male influencees

In terms of their overall remote influence effects there appeared to be nodifference between experimenterinfluencers in Study 1 (r = +06 - 05 and

- 06) In Study 2 effect size measures indicated that the experimenters dif-fered in remote influence effect sizes (r = +17 and +30) Due to the smaller Nthis difference was not found to be statistically significant Post hoc combin-ing Studies 1 and 2 for experimenters JW and CW (ZMcD was not an experi-menter in Study 2) the effect sizes are r = +08 and r = +20 respectively

We conducted a post hoc analysis of the effects of influenceeinfluencer sexpairings on remote influence scores Our two studies came up with contradic-tory findings each statistically significant Study one found that mixed sexpairings had significantly higher remote influence scores than same sex pair-ings This replicated the post hoc finding of Schlitz and LaBerge (1997) in con-nection with the autonomic detection of remote observation Study 2 foundthat same sex pairings had significantly higher remote influence scores thanmixed sex pairings Our contradictory findings do not help to clarify the pic-ture on the importance or otherwise of sex pairings It is quite possible that sexis not the most salient aspect of the dynamic between pairs Age personalityand familiarity are likely to be important variables moderating the relationshipbetween influencer and influencee

There are many more variables that we could look at but as the number ofanalyses multiplies so does the opportunity for spurious significant resultsFor this reason we have named only two formal hypotheses for each experi-ment All other analyses are regarded as exploratory and any trends that wehave highlighted above would require replication

In conclusion these two studies have found no evidence of a difference inremote influence effects when influencees are asked to adopt cooperative men-tal strategies compared to when they are asked to adopt blocking mental strate-gies In fact contrary to expectation both studies found slightly higher remoteinfluence effects for the blocking condition compared to the co-operating con-dition On this basis we have found no indication of limiting conditions for re-mote influence effects Therefore while anecdotal advice exists on how to de-fend oneself against a remote physiological influence attempt in the realworld laboratory evidence to date has provided no support for this advice

Given the importance of this question both in practical terms and in theoret-ical terms we argue that this line of research deserves continued investigationIf a way could be found to conduct research with a more realistic scenario mdash where the remote influence was not just neutrally and abstractlydefined as unwelcome mdash then our research might have more meaningful thingsto say about the parameters of remote influence It is interesting and perhapssurprising to note that when Watt has lectured on this topic to informed lay au-diences the feedback has often been that researchers must use more potential-ly harmful or disruptive remote influence (of course with the informed consentof the participants) While there are clear ethical inhibitions when consideringpossible physiological influence perhaps research using higher level influencewould be more acceptable For example one could attempt to disrupt or en-hance the concentration of chess players who were aware of the influence

Blocking vs Co-operating 533

534 C Watt et al

attempt and who were concurrently blocking or co-operating with the influ-ence attempt Research already exists to suggest that remote influence does notjust operate at physiological levels Attention focusing for example can be fa-cilitated by a remote helper (Brady amp Morris 1997 Braud et al 1995) Futureresearch in this area must therefore expand beyond physiological influences

Acknowledgements

The authors gratefully acknowledge the financial support they receivedfrom the Institut fuumlr Grenzgebiete der Psychologie und PsychohygieneFreiburg Germany while conducting these studies We would also like tothank Dr Paul Stevens for generously helping to solve software difficultieswhen preparing for these studies and the anonymous reviewer for helpfulcomments Address reprint requests to Dr C Watt Department of PsychologyUniversity of Edinburgh 7 George Square Edinburgh EH8 9JZ Scotland UK

ReferencesBrady C amp Morris R L (1997) Attention focusing facilitated through remote mental interac-

tion A replication and exploration of parameters In R Wiseman (Ed) Proceedings of Pre-sented Papers The Parapsychological Association 40th Annual Convention held in conjunc-tion with The Society for Psychical Research pp 73ndash91

Braud W G (1985) Blockingshielding psychic functioning through psychological and psychictechniques A report of three preliminary studies In RA White and J Solfvin (Eds) Re-search in Parapsychology 1984 Metuchen NJ Scarecrow Press pp 42ndash44

Braud W G amp Schlitz M J (1991) Consciousness interactions with remote biological systemsAnomalous intentionality effects Subtle Energies 2 1

Braud W G Schlitz M J Collins J and Klitch H (1985) Further studies of the bio-PK effectFeedback blocking specificitygenerality In RA White amp J Solfvin (Eds) Research inParapsychology 1984 Metuchen NJ Scarecrow Press pp 45ndash48

Braud W G Shafer D McNeill K and Guerra V (1995) Attention focusing facilitatedthrough remote mental interaction Journal of the American Society for Psychical Research89 103

Cohen J (1997) Statistical Power Analysis for the Behavioral Sciences New York AcademicPress

Dalton K (1994) Remotely influenced ESP performance in a computer task A preliminarystudy Proceedings of the 37th Annual Convention of the Parapsychological Association pp95ndash103

Dossey L (1994) Healing and the mind Is there a dark side Journal of Scientific Exploration 873

Halifax-Grof J (1974) Hex death In A Angoff amp D Barth (Eds) Parapsychology and Anthro-pology Proceedings of an International Conference Held in London England August 29ndash311973 NY Parapsychology Foundation

Hope M (1983) Practical Techniques of Psychic Self-Defence Wellingborough Northampton-shire Aquarian Press

Marks DF (1973) Visual imagery differences in the recall of pictures British Journal of Psy-chology 64 17

Nash CB (1982) Psychokinetic control of bacterial growth Journal of the Society for PsychicalResearch 51 217

Radin DI Taylor RK and Braud WG (1995) Remote mental influence of human electroder-mal activity A pilot replication European Journal of Parapsychology 11 19

Schlitz M amp Braud W G (1997) Distant intentionality and healing Assessing the evidence Al-ternative Therapies 3 62

Schlitz M J amp LaBerge S (1997) Covert observation increases skin conductance in subjectsunaware of when they are being observed A replication Journal of Parapsychology 61 185

Schmeidler G R (1958) Agent-percipient relationships Journal of the American Society forPsychical Research 52 47

Schmeidler G R (1961) Evidence for two kinds of telepathy International Journal of Parapsy-chology 3 5

Tart C T (1979) A survey of expert opinion on potentially negative uses of psi United Statesgovernment interest in psi and the level of research funding of the field In WG Roll (Ed) Re-search in Parapsychology 1978 Metuchen NJ Scarecrow Press pp 54

Tart C T amp Labore C M (1986) Attitudes towards strongly functioning psi A preliminary sur-vey Journal of the America Society for Psychical Research 80 163

Tart C T (1986) Psychicsrsquo fears of psychic powers Journal of the American Society for Psychi-cal Research 80 278

Watt C A McDermott Z amp Ravenscroft J (1997) Exploring the limits of remote mental inter-action A comparison of blocking versus co-operating strategies In R Wiseman (Ed) Pro-ceedings of the 40th Annual Convention of the Parapsychological Association held in conjunc-tion with the Society for Psychical Research Brighton England pp 464ndash476

Blocking vs Co-operating 535

Incomplete Data

One participant declined to give their age on the PIF form so the descriptivestatistics on age do not include this participant (so N = 31) Also due to a tech-nical hitch the exploratory variance scores were lost for one participant so N = 31 for this measure There were no drop-outs from this study

Remote Influence Measure

Table 1 presents a summary of the remote influence data in the form ofmean Percentage Score Index for co-operating and blocking conditions Therewas no significant difference between PSI for co-operating and blocking con-ditions (t = - 202 df = 31) Therefore Hypothesis 1 (predicting significantlygreater PSI scores for co-operating condition compared to blocking condition)was not supported by these results As planned the PSI scores for co-operatingand blocking runs were combined to examine Hypothesis 2 (of an overall re-mote mental influence effect in line with the calming and activating intentionsof the influencee) Mean PSI scores were exactly at chance expectation (t = - 031 df = 31) therefore there was no support for Hypothesis 2

Exploratory Questions

Variance Measure

Table 2 presents summary data for the exploratory measure of the amount ofvariance shown in each condition (lt 50 represents reduced variance gt 50 rep-resents increased variance) There was no significant difference between vari-ance scores for co-operating and blocking conditions (t = 835 df = 30) andwhen the conditions were combined variance did not depart from chance ex-pectation (t = - 244 df = 30) It is interesting to note however that the vari-ance measure gives a much wider range of scores than the PSI measure

Blocking vs Co-operating 523

TABLE 1Summary Data Mean PSI Scores

(N = 32 MCE = 50 range in brackets)

Co-operate PSI Block PSI Combined PSI

4996 5003 5000(4692- 5215) (4653- 5404) (4853- 5309)

TABLE 2Summary Data Mean Variance Scores(N = 31 MCE = 50 range in brackets)

Co-operate Variance Block Variance Combined Variance

5089 4864 4977(2648- 7012) (2492- 6484) (4116- 6297)

524 C Watt et al

suggesting that the variance measure may potentially be a more sensitive indi-cator of remote influence effect than PSI

Vividness of Mental Imagery

Low VVIQ scores represent participants reporting highly vivid imagery andhigh VVIQ scores represent those reporting little or no imagery The meanscore was 4275 (range from 21 to 80) Using a Pearson correlation we foundthat there was no evidence to suggest that those participants with low VVIQscores (high imagery) were better blockers than those with high scores (VVIQvs PSI blocking r = - 113) We also found that there was no evidence of a cor-relation between vividness of imagery and overall remote influence effectsusing the combined Percentage Score Index measure (VVIQ vs PSI overall r =- 077)

Influenceesrsquo Blocking and Co-operating Strategies

Influencees completed a questionnaire (after the session was over) askingthem to describe what strategies they used for blocking and co-operating Cat-egorizing these responses we found a difference in the kinds of strategies usedwhen blocking compared to co-operating as shown in Table 3 In the blockingcondition the majority of participants (469) used visual imagery strategiesFewer (281) participants used a combination of imagery and cognitivestrategies (such as repeating words related to blocking) while 25 reportedusing purely cognitive strategies for blocking (such as adopting a resistant anduncooperative attitude or thinking of irrelevancies) In the co-operating con-dition fewer participants used visual imagery alone (333) The majority ofparticipants (424) reported that their strategies for co-operating includedadopting a feeling of passiveness openness or relaxation whereas no partici-pants adopted this kind of strategy for the blocking condition 121 of partic-ipants used cognitive strategies alone (such as adopting a co-operative atti-tude) and the same proportion used a combination of visual imagery andcognitive strategies

Therefore there was a tendency for participants to use visual imagery forblocking and relaxationpassivity or visual imagery for co-operating This wasdespite the fact that we asked participants to attempt to adopt similar levels ofactivation for both co-operating and blocking strategies There was also a ten-dency not surprisingly for those participants with low VVIQ scores (ie vivid

TABLE 3Influenceesrsquo Mental Strategies ()

Vis Im Cognitive VisCog OpenRelax

Blocking 469 250 281 00Co-operating 333 121 121 424

imagers) to use visual imagery strategies Those who had high VVIQ scoresalso used visual imagery but more often used this in combination with othertactics

Other Qualitative Reports from Participants According to their post-ses-sion questionnaire responses participants generally felt comfortable about theattempt to be influenced (N = 24) Three participants said they felt apprehen-sive or uncomfortable with the idea of being influenced Interestingly of thesethree two were actively involved with parapsychology and were aware of theharmful psi implications discussed at the beginning of this report The thirdwas a psychology student who had attended a lecture given by Watt in whichthe possibilities of harmful psi were discussed Another group of three partici-pants said they would be uncomfortable with the idea of being influenced ifthey believed such influence was possible

Seven participants said they found it was difficult to sustain concentrationfor the 10 minutes that comprised each condition Five participants found thescreensaver distracting or disturbing Two participants specifically mentionedthat they found blocking a more taxing task than co-operating

Sex Effects

There were more males (N = 20) than females (N = 12) in this study howev-er there was no significant difference in PSI scores for remote physiology in-fluence between males and females (male PSI vs female PSI t = 1085 male PSI = 4987 female PSI = 5021)

Experimenter Effects

We studied whether there was any difference in PSI scores obtained by thethree experimenters The results are shown in Table 4 There seems to be nosign of experimenter effects in study one

InfluenceeInfluencer Sex Pairings

Table 5 gives the mean PSI scores for different influenceeinfluencer pair-ings Taking the combined PSI scores it can be seen that the highest remote in-fluence effect is for FM pairs followed by MF pairs followed by FF pairsand the lowest scoring is for MM pairs Post hoc it was found that the mixedsex pairings showed significantly higher remote influence scores than the same

Blocking vs Co-operating 525

TABLE 4Remote Influence Results by Experimenter

Experimenter N sessions Coop PSI Block PSI Combined PSI

CW 12 4983 5011 4997ZMcD 10 5015 4998 5006JR 10 4994 4999 4996

526 C Watt et al

sex pairings (t = - 2016 df = 30 p = 053 2 - t) This replicates the effect foundby Schlitz and LaBerge (1997)

Study One Conclusions

In Study 1 we developed a methodology to examine intentional resistance toa remote mental influence attempt (Watt et al 1997) We contrasted influ-enceesrsquo physiological activity when they were mentally co-operating with aremote attempt to influence their physiology with physiological activity wheninfluencees were mentally blocking the influence attempt We found no evi-dence for any remote influence effect overall and no difference betweenblocking and co-operating conditions Our qualitative measures yielded someinteresting information about participantsrsquo mental strategies and reactions dur-ing the remote influence session We found no difference between experi-menters in remote influence scores but we did find that mixed sex influ-enceeinfluencer pairings scored more highly than same sex pairings

Study Two Method

Study 2 was planned to explore in more depth the nature of the relationshipbetween aspects of personality and cognition and psi performance Partici-pants took part in three testing sessions approximately one week apart Thefirst consisted of various psychological measures the second consisted of aforced choice ESP task and the third consisted of a replication of the blocking-co-operating study (Study 1) described above The psychological measureswere not scored until after completion of the remote influence session Partici-pants received feedback about their forced choice ESP performance at the con-clusion of the second session For ease of interpretation the present paper isrestricted to a description of the third testing session which was essentially areplication of study one with fewer influencers and greater statistical power

We introduced some slight changes to the methodology of Study 2 based onour experiences with Study 1 At the start of each of the two runs in Study 1the participant was given auditory instructions either to ldquoblock nowrdquo or ldquoco-operate nowrdquo We observed that the auditory instructions appeared to affect theparticipantrsquos EDA as if they were startled As planned the first recordingepoch in each run was not used in analysis to give the participantrsquos EDA time

TABLE 5Mean PSI Scores for Different InfluenceeInfluencer Sex Pairings

Pairing N Coop PSI Block PSI Combined PSI

M-F 6 4932 5069 5001F-M 6 5043 5034 5048F-F 6 5022 4952 4993M-M 14 4988 4975 4981

to settle down However in order to remove any remaining potential artifactu-al effects on EDA in Study 2 only visual instructions were given

Another difference in methodology concerned the screensaver viewed bythe participants during the experimental session In Study 1 some participantsreported that the series of patterns produced by the computer were being dis-played at too fast a rate such that some participants found it distracting or dis-turbing Again we felt that this may have an artifactual effect on EDA so wedecided for study two to reduce the speed of the screensaver

Thirdly in Study 2 we did not give participants detailed information abouttheir remote influence performance until the feedback letters that were distrib-uted at the conclusion of the entire study The reason for this change was mere-ly to streamline the procedure on what was already a time-consuming experi-ment Finally in another streamlining change we did not take VVIQ measuresin Study 2

Planned Hypotheses and Analyses

These were the same as with Study 1 That is hypothesis 1 predicted higherPSI scores for the co-operating condition than for the blocking condition(measured with a related t-test one-tailed) If no significant difference wasfound between blocking PSI and co-operating PSI scores the two would becombined to look at overall remote influence effect (hypothesis 2 single-meant-test one-tailed) If a significant difference was found with hypothesis 1 theanalysis for hypothesis 2 would be conducted with the data for co-operatingsessions only

Exploratory Questions

As with study one we took an exploratory variance measure of remote EDAinfluence We also gathered qualitative descriptions of the mental strategiesparticipants adopted whilst blocking and co-operating We also looked at sexeffects experimenter effects and post hoc influenceeinfluencer sex pairings

Participants

Study 2 had two experimenterinfluencers (Watt amp Ravenscroft) After aparticipant was recruited (usually by Ravenscroft) they were assigned to oneof the two experimenters for all three testing sessions This assignment wasdone unsystematically according to the preferences and availability of the ex-perimenters As planned 50 volunteer participants were drawn from the gener-al participant pool at the Koestler Parapsychology Unit and by word ofmouth All participants completed the Participant Information Form and aConsent form prior to the study The participants consisted of 33 females and17 males Average age was 41 years range from 22 years to 76 years of age Inorder to increase the studyrsquos relevance to the general population undergradu-ate psychology students were not recruited as participants The main criterion

Blocking vs Co-operating 527

528 C Watt et al

for participation was that the participants were interested in and open-mindedabout parapsychology Again there were no drop-outs from this study Allparticipants received a detailed personal feedback letter at the conclusion ofthe study

Study Two Results and Discussion

Remote Influence Measure

PSI scores are summarized in Table 6 which shows that the mean PSI scorewhen co-operating was nonsignificantly lower than mean PSI score whenblocking (t = - 595 df = 49) Hypothesis 1 was therefore not supported Asplanned in the absence of any significant difference between blocking and co-operating PSI scores the two conditions were combined to test hypothesis 2which predicted overall significant remote mental influence performance Thecombined PSI score was significantly greater than mean chance expectation of50 (t = 1806 df = 49 p = 038 1 - t) Therefore hypothesis 2 was supportedgiving evidence of an overall remote influence effect upon psychophysiologyin the prespecified direction

The failure to support hypothesis 1 confirms what we found in Study 1 thatthere was no significant difference between blocking and co-operating condi-tions on remote influence scores Indeed both studies found a slight effect inthe opposite direction to that predicted that is slightly higher PSI scores whenblocking was the influenceersquos intention than when co-operating was their in-tention It was encouraging to find significant support for a remote influenceeffect in the predicted direction for blocking and co-operating conditions com-bined No such trend was found in the previous study though with a smallernumber of participants (N = 32) the earlier study had lower statistical powerthan the present study

Exploratory Questions

Variance Measure

Table 7 shows the results for the exploratory measure of EDA variance lt 50represents reduced variance gt 50 represents increased variance We found nodifference between variance scores for cooperate versus block conditions(t = 150 df = 49) When these are combined the resulting figure does not de-viate from chance expectation (t = 440 df = 49) Thus as in Study 1 it seems

TABLE 6Summary Data Mean PSI Scores

(N = 50 MCE = 50 range in brackets)

Co-operate PSI Block PSI Combined PSI

5008 5019 5016(4821- 5221) (4807- 5557) (4900- 5296)

that the exploratory variance measure does not reveal any additional indica-tion of the operation of a remote mental influence in this study However thewider range of variance scoring continues to suggest that this measure mayhave greater potential sensitivity than the summary PSI score

Influenceesrsquo Blocking and Co-operating Strategies

We were interested in learning more about what particular strategies partici-pants chose to adopt when blocking and co-operating As in Study 1 a post-session questionnaire was used to gather this information Table 8 shows thepercentage of participants who adopted each kind of strategy

The table shows a very similar pattern to that found in Study 1 The mainpoint of interest is that almost no participants took a passive approach toblocking (ie adopting an openrelaxing strategy) A much higher proportionpreferred to adopt this approach when co-operating The preferred mode forboth conditions was to use some form of visual imagery such as imagining ashield or a screen when blocking and imagining an open channel such as ariver when co-operating

The majority of participants reported no concerns over the possibility ofhaving their physiology remotely influenced by another person as found instudy one

Sex Effects

As in study one the sexes were not equally represented among participantsWhereas Study 1 had an excess of male participants in Study 2 twice as manyfemales (N = 33) as males (N = 17) participated It was therefore important tolook for sex differences since any such differences could confound the inter-pretation of overall trends We found no difference between sexes when co-op-erating (female mean PSI score = 5008 male mean PSI score = 5006 t = 107 df = 48) There was a nonsignificant trend for females to have higher

Blocking vs Co-operating 529

TABLE 7Summary Data Mean Variance Scores(N = 50 MCE = 50 range in brackets)

Co-operate Variance Block Variance Combined Variance

5076 5054 5034(3164- 7279) (3572- 6813) (3749- 6671)

TABLE 8Influenceesrsquo Mental Strategies ()

Visual Imagery Cognitive VisCog OpenRelax

Blocking 64 26 6 2Co-operating 60 8 12 20

530 C Watt et al

PSI scores than males when blocking (mean PSI scores 5030 and 4998 re-spectively t = 992 df = 48) Combining blocking and co-operating conditionsthere was no overall difference between the sexes (mean PSI scores 5020 forfemales 5008 for males t = 665 df = 48)

Experimenter Effects

Study 2 found a statistically significant overall remote influence effect(combined PSI = 50155 t = 1806) Table 9 shows the breakdown of remoteinfluence scores for the two experimenters CW and JR Looking at the overalldata for the two experimenters we see that CWrsquos sessions were associatedwith most of the remote influence effect However due to the smaller N CWrsquossessions are not independently significant (t = 1632) and there is not a signifi-cant difference between CW and JRrsquos overall PSI scores (t = 678) Post hocwe calculated effect size r according to the formula r = [t2(t2+df)]12 for the twoexperimenters For JR (t = 817 df = 22) we found a small effect sizer = +17 For CW (t = 1632 df = 26) the effect size was moderate r = +30This may be some post hoc indication of an experimenter effect in Study 2

InfluenceeInfluencer Sex Pairings

As can be seen from Table 10 Study 2 found exactly the opposite trend tothat found in Study 1 On a post hoc test same sex pairings scored significant-ly higher on the remote influence PSI measure than mixed sex pairings(t = - 2034 df = 48 p = 048 2 - t) This contradicts Schlitz and LaBergersquos(1997) finding of greater autonomic detection of remote staring for mixed sexpairs

TABLE 9Remote Influence Results by Experimenter

Experimenter N sessions Coop PSI Block PSI Combined PSI

CW 27 5011 5026 5021JR 23 5003 5010 5009

TABLE 10Mean PSI Scores for Different InfluenceeInfluencer Sex Pairings

Pairing N Coop PSI Block PSI Combined PSI

MF 11 4977 5009 4997FM 17 4983 5021 5003FF 16 5035 5038 5038MM 6 5060 4978 5027

General Summary and Conclusions

We have described the findings of two studies exploring the limits of remotemental influence effects We wished to see whether the mental strategiesadopted by the influencee would affect remote influence scores In particularcould an influencee block or resist a remote influence attempt We recorded in-fluenceesrsquo electrodermal activity while the influencer attempted to calm to ac-tivate or to have no influence over the influenceesrsquo physiology In each studythere were two conditions reflecting the influenceesrsquo designated mental strate-gies in response to the remote influence attempt blocking and co-operating

There were only two formal hypotheses 1) that remote influence scoreswould be significantly higher during the co-operating condition than duringthe blocking condition and 2) that overall there would be significant evidenceof remote mental influence on physiology in the designated direction (iecalming when calming was the influencerrsquos intention activating when activat-ing was the influencerrsquos intention) Contrary to prediction Study 1 found thatremote influence effects were higher in the block condition than in the co-op-erate condition (but not to a significant extent) Also overall remote influenceperformance was at chance level This study therefore found no support for ei-ther of the formal hypotheses Study 2 consisted of a replication of Study 1with a few minor adjustments plus additional measures that were not reportedin this paper Again contrary to prediction remote influence scores were non-significantly higher in the block condition than in the co-operate conditionThis replicates the trend found in Study 1 failing to support hypothesis oneThese results suggest that the mental strategy adopted by the influencee in theface of a remote influence attempt has no effect on remote influence perfor-mance Referring back to the questions posed in the introduction then wehave not found any evidence to suggest that it is possible to resist a remote in-fluence effect and there is as yet no indication of the possible limiting condi-tions for remote mental influence

Unlike Study 1 Study 2 found statistically significant evidence of an overallremote influence effect in the prespecified direction (ie calming of EDAwhen that was the influencerrsquos intention activation of EDA when that was theinfluencerrsquos intention) In order to facilitate comparison with a recent meta-an-alytic review of remote influence studies we calculated the effect size (r) ac-cording to the formula r = [t2(t2+df)]12 For their review of 19 EDA influencestudies Schlitz and Braud (1997) found an average effect size of r = +25 OurStudy 1 with overall t = - 031 df = 31 gives a negligible effect size of r = - 006 Study 2 with overall t = 1806 df = 49 gives r = +25 Study 2 there-fore replicates the mean effect size found in EDA studies conducted by otherresearchers in other parapsychology laboratories Post hoc we can combinethe data of the two studies for greater statistical power Overall t = 1176 df = 81 giving a small effect size of r = +13

We can find no clear reason why Study 2 found an EDA influence effectwhile Study 1 did not A number of factors are worth considering 1) The

Blocking vs Co-operating 531

532 C Watt et al

significant result could be due to chance or undetected artifact 2) Study 2 im-plemented what we considered to be minor methodological improvements thatmight have removed some artifactual fluctuations from the EDA data that weobserved in study one 3) Each experimenter in Study 2 conducted three testingsessions with their participant By the third session in which the remote influ-ence measure was taken experimenters were on quite friendly terms with theirparticipants perhaps more so than in Study 1 where each participant came infor only one testing session 4) The participants in Study 2 were more represen-tative of the general population we felt than those in Study 1 Study 1 in-volved more undergraduate psychology students and associates of the experi-menters 5) The experimenters in study one were less experienced with theremote influence protocol than when they conducted Study 2

In addition to the formal hypotheses we reported a number of exploratorymeasures Study 1 found that influenceesrsquo self-reported vividness of visual im-agery appeared to be unrelated to the remote influence performance In bothstudies influenceesrsquo qualitative descriptions of the actual mental strategiesthey employed when blocking and co-operating revealed a tendency for differ-ent strategies to be preferred for the different conditions The use of visual im-agery was popular for both blocking (eg visualizing being surrounded by aprotective cocoon) and co-operating (eg visualizing a river flowing betweeninfluencer and influencee) However a considerable proportion of participantsreported adopting an open or relaxed frame of mind when co-operating whilstalmost no participants chose this strategy when blocking This was despite thefact that prior to each session we encouraged participants to adopt comparablestrategies for blocking and co-operating

It was interesting and perhaps surprising to note that in both studies themajority of participants reported no feelings of concern or discomfort at thepossibility of having their physiology remotely influenced Those who wereconcerned tended to be those who knew more about parapsychology and thepossible harmful applications of remote influence A small proportion of par-ticipants was unconcerned because they did not believe such an influence waspossible in the first place

Both studies employed an exploratory variance measure of EDA influenceThis measure showed a greater range of scoring overall thus perhaps greaterpotential sensitivity to a remote influence effect However the variance mea-sure showed no indication of a remote influence effect To date researchershave employed a variety of indicators of remote influence as the literaturegrows progress may be made on identifying which is the most appropriate in-dicator

Neither study found significant sex differences in remote influence effectsthough in both studies female influencees tended to have higher remote influ-ence scores than male influencees

In terms of their overall remote influence effects there appeared to be nodifference between experimenterinfluencers in Study 1 (r = +06 - 05 and

- 06) In Study 2 effect size measures indicated that the experimenters dif-fered in remote influence effect sizes (r = +17 and +30) Due to the smaller Nthis difference was not found to be statistically significant Post hoc combin-ing Studies 1 and 2 for experimenters JW and CW (ZMcD was not an experi-menter in Study 2) the effect sizes are r = +08 and r = +20 respectively

We conducted a post hoc analysis of the effects of influenceeinfluencer sexpairings on remote influence scores Our two studies came up with contradic-tory findings each statistically significant Study one found that mixed sexpairings had significantly higher remote influence scores than same sex pair-ings This replicated the post hoc finding of Schlitz and LaBerge (1997) in con-nection with the autonomic detection of remote observation Study 2 foundthat same sex pairings had significantly higher remote influence scores thanmixed sex pairings Our contradictory findings do not help to clarify the pic-ture on the importance or otherwise of sex pairings It is quite possible that sexis not the most salient aspect of the dynamic between pairs Age personalityand familiarity are likely to be important variables moderating the relationshipbetween influencer and influencee

There are many more variables that we could look at but as the number ofanalyses multiplies so does the opportunity for spurious significant resultsFor this reason we have named only two formal hypotheses for each experi-ment All other analyses are regarded as exploratory and any trends that wehave highlighted above would require replication

In conclusion these two studies have found no evidence of a difference inremote influence effects when influencees are asked to adopt cooperative men-tal strategies compared to when they are asked to adopt blocking mental strate-gies In fact contrary to expectation both studies found slightly higher remoteinfluence effects for the blocking condition compared to the co-operating con-dition On this basis we have found no indication of limiting conditions for re-mote influence effects Therefore while anecdotal advice exists on how to de-fend oneself against a remote physiological influence attempt in the realworld laboratory evidence to date has provided no support for this advice

Given the importance of this question both in practical terms and in theoret-ical terms we argue that this line of research deserves continued investigationIf a way could be found to conduct research with a more realistic scenario mdash where the remote influence was not just neutrally and abstractlydefined as unwelcome mdash then our research might have more meaningful thingsto say about the parameters of remote influence It is interesting and perhapssurprising to note that when Watt has lectured on this topic to informed lay au-diences the feedback has often been that researchers must use more potential-ly harmful or disruptive remote influence (of course with the informed consentof the participants) While there are clear ethical inhibitions when consideringpossible physiological influence perhaps research using higher level influencewould be more acceptable For example one could attempt to disrupt or en-hance the concentration of chess players who were aware of the influence

Blocking vs Co-operating 533

534 C Watt et al

attempt and who were concurrently blocking or co-operating with the influ-ence attempt Research already exists to suggest that remote influence does notjust operate at physiological levels Attention focusing for example can be fa-cilitated by a remote helper (Brady amp Morris 1997 Braud et al 1995) Futureresearch in this area must therefore expand beyond physiological influences

Acknowledgements

The authors gratefully acknowledge the financial support they receivedfrom the Institut fuumlr Grenzgebiete der Psychologie und PsychohygieneFreiburg Germany while conducting these studies We would also like tothank Dr Paul Stevens for generously helping to solve software difficultieswhen preparing for these studies and the anonymous reviewer for helpfulcomments Address reprint requests to Dr C Watt Department of PsychologyUniversity of Edinburgh 7 George Square Edinburgh EH8 9JZ Scotland UK

ReferencesBrady C amp Morris R L (1997) Attention focusing facilitated through remote mental interac-

tion A replication and exploration of parameters In R Wiseman (Ed) Proceedings of Pre-sented Papers The Parapsychological Association 40th Annual Convention held in conjunc-tion with The Society for Psychical Research pp 73ndash91

Braud W G (1985) Blockingshielding psychic functioning through psychological and psychictechniques A report of three preliminary studies In RA White and J Solfvin (Eds) Re-search in Parapsychology 1984 Metuchen NJ Scarecrow Press pp 42ndash44

Braud W G amp Schlitz M J (1991) Consciousness interactions with remote biological systemsAnomalous intentionality effects Subtle Energies 2 1

Braud W G Schlitz M J Collins J and Klitch H (1985) Further studies of the bio-PK effectFeedback blocking specificitygenerality In RA White amp J Solfvin (Eds) Research inParapsychology 1984 Metuchen NJ Scarecrow Press pp 45ndash48

Braud W G Shafer D McNeill K and Guerra V (1995) Attention focusing facilitatedthrough remote mental interaction Journal of the American Society for Psychical Research89 103

Cohen J (1997) Statistical Power Analysis for the Behavioral Sciences New York AcademicPress

Dalton K (1994) Remotely influenced ESP performance in a computer task A preliminarystudy Proceedings of the 37th Annual Convention of the Parapsychological Association pp95ndash103

Dossey L (1994) Healing and the mind Is there a dark side Journal of Scientific Exploration 873

Halifax-Grof J (1974) Hex death In A Angoff amp D Barth (Eds) Parapsychology and Anthro-pology Proceedings of an International Conference Held in London England August 29ndash311973 NY Parapsychology Foundation

Hope M (1983) Practical Techniques of Psychic Self-Defence Wellingborough Northampton-shire Aquarian Press

Marks DF (1973) Visual imagery differences in the recall of pictures British Journal of Psy-chology 64 17

Nash CB (1982) Psychokinetic control of bacterial growth Journal of the Society for PsychicalResearch 51 217

Radin DI Taylor RK and Braud WG (1995) Remote mental influence of human electroder-mal activity A pilot replication European Journal of Parapsychology 11 19

Schlitz M amp Braud W G (1997) Distant intentionality and healing Assessing the evidence Al-ternative Therapies 3 62

Schlitz M J amp LaBerge S (1997) Covert observation increases skin conductance in subjectsunaware of when they are being observed A replication Journal of Parapsychology 61 185

Schmeidler G R (1958) Agent-percipient relationships Journal of the American Society forPsychical Research 52 47

Schmeidler G R (1961) Evidence for two kinds of telepathy International Journal of Parapsy-chology 3 5

Tart C T (1979) A survey of expert opinion on potentially negative uses of psi United Statesgovernment interest in psi and the level of research funding of the field In WG Roll (Ed) Re-search in Parapsychology 1978 Metuchen NJ Scarecrow Press pp 54

Tart C T amp Labore C M (1986) Attitudes towards strongly functioning psi A preliminary sur-vey Journal of the America Society for Psychical Research 80 163

Tart C T (1986) Psychicsrsquo fears of psychic powers Journal of the American Society for Psychi-cal Research 80 278

Watt C A McDermott Z amp Ravenscroft J (1997) Exploring the limits of remote mental inter-action A comparison of blocking versus co-operating strategies In R Wiseman (Ed) Pro-ceedings of the 40th Annual Convention of the Parapsychological Association held in conjunc-tion with the Society for Psychical Research Brighton England pp 464ndash476

Blocking vs Co-operating 535

524 C Watt et al

suggesting that the variance measure may potentially be a more sensitive indi-cator of remote influence effect than PSI

Vividness of Mental Imagery

Low VVIQ scores represent participants reporting highly vivid imagery andhigh VVIQ scores represent those reporting little or no imagery The meanscore was 4275 (range from 21 to 80) Using a Pearson correlation we foundthat there was no evidence to suggest that those participants with low VVIQscores (high imagery) were better blockers than those with high scores (VVIQvs PSI blocking r = - 113) We also found that there was no evidence of a cor-relation between vividness of imagery and overall remote influence effectsusing the combined Percentage Score Index measure (VVIQ vs PSI overall r =- 077)

Influenceesrsquo Blocking and Co-operating Strategies

Influencees completed a questionnaire (after the session was over) askingthem to describe what strategies they used for blocking and co-operating Cat-egorizing these responses we found a difference in the kinds of strategies usedwhen blocking compared to co-operating as shown in Table 3 In the blockingcondition the majority of participants (469) used visual imagery strategiesFewer (281) participants used a combination of imagery and cognitivestrategies (such as repeating words related to blocking) while 25 reportedusing purely cognitive strategies for blocking (such as adopting a resistant anduncooperative attitude or thinking of irrelevancies) In the co-operating con-dition fewer participants used visual imagery alone (333) The majority ofparticipants (424) reported that their strategies for co-operating includedadopting a feeling of passiveness openness or relaxation whereas no partici-pants adopted this kind of strategy for the blocking condition 121 of partic-ipants used cognitive strategies alone (such as adopting a co-operative atti-tude) and the same proportion used a combination of visual imagery andcognitive strategies

Therefore there was a tendency for participants to use visual imagery forblocking and relaxationpassivity or visual imagery for co-operating This wasdespite the fact that we asked participants to attempt to adopt similar levels ofactivation for both co-operating and blocking strategies There was also a ten-dency not surprisingly for those participants with low VVIQ scores (ie vivid

TABLE 3Influenceesrsquo Mental Strategies ()

Vis Im Cognitive VisCog OpenRelax

Blocking 469 250 281 00Co-operating 333 121 121 424

imagers) to use visual imagery strategies Those who had high VVIQ scoresalso used visual imagery but more often used this in combination with othertactics

Other Qualitative Reports from Participants According to their post-ses-sion questionnaire responses participants generally felt comfortable about theattempt to be influenced (N = 24) Three participants said they felt apprehen-sive or uncomfortable with the idea of being influenced Interestingly of thesethree two were actively involved with parapsychology and were aware of theharmful psi implications discussed at the beginning of this report The thirdwas a psychology student who had attended a lecture given by Watt in whichthe possibilities of harmful psi were discussed Another group of three partici-pants said they would be uncomfortable with the idea of being influenced ifthey believed such influence was possible

Seven participants said they found it was difficult to sustain concentrationfor the 10 minutes that comprised each condition Five participants found thescreensaver distracting or disturbing Two participants specifically mentionedthat they found blocking a more taxing task than co-operating

Sex Effects

There were more males (N = 20) than females (N = 12) in this study howev-er there was no significant difference in PSI scores for remote physiology in-fluence between males and females (male PSI vs female PSI t = 1085 male PSI = 4987 female PSI = 5021)

Experimenter Effects

We studied whether there was any difference in PSI scores obtained by thethree experimenters The results are shown in Table 4 There seems to be nosign of experimenter effects in study one

InfluenceeInfluencer Sex Pairings

Table 5 gives the mean PSI scores for different influenceeinfluencer pair-ings Taking the combined PSI scores it can be seen that the highest remote in-fluence effect is for FM pairs followed by MF pairs followed by FF pairsand the lowest scoring is for MM pairs Post hoc it was found that the mixedsex pairings showed significantly higher remote influence scores than the same

Blocking vs Co-operating 525

TABLE 4Remote Influence Results by Experimenter

Experimenter N sessions Coop PSI Block PSI Combined PSI

CW 12 4983 5011 4997ZMcD 10 5015 4998 5006JR 10 4994 4999 4996

526 C Watt et al

sex pairings (t = - 2016 df = 30 p = 053 2 - t) This replicates the effect foundby Schlitz and LaBerge (1997)

Study One Conclusions

In Study 1 we developed a methodology to examine intentional resistance toa remote mental influence attempt (Watt et al 1997) We contrasted influ-enceesrsquo physiological activity when they were mentally co-operating with aremote attempt to influence their physiology with physiological activity wheninfluencees were mentally blocking the influence attempt We found no evi-dence for any remote influence effect overall and no difference betweenblocking and co-operating conditions Our qualitative measures yielded someinteresting information about participantsrsquo mental strategies and reactions dur-ing the remote influence session We found no difference between experi-menters in remote influence scores but we did find that mixed sex influ-enceeinfluencer pairings scored more highly than same sex pairings

Study Two Method

Study 2 was planned to explore in more depth the nature of the relationshipbetween aspects of personality and cognition and psi performance Partici-pants took part in three testing sessions approximately one week apart Thefirst consisted of various psychological measures the second consisted of aforced choice ESP task and the third consisted of a replication of the blocking-co-operating study (Study 1) described above The psychological measureswere not scored until after completion of the remote influence session Partici-pants received feedback about their forced choice ESP performance at the con-clusion of the second session For ease of interpretation the present paper isrestricted to a description of the third testing session which was essentially areplication of study one with fewer influencers and greater statistical power

We introduced some slight changes to the methodology of Study 2 based onour experiences with Study 1 At the start of each of the two runs in Study 1the participant was given auditory instructions either to ldquoblock nowrdquo or ldquoco-operate nowrdquo We observed that the auditory instructions appeared to affect theparticipantrsquos EDA as if they were startled As planned the first recordingepoch in each run was not used in analysis to give the participantrsquos EDA time

TABLE 5Mean PSI Scores for Different InfluenceeInfluencer Sex Pairings

Pairing N Coop PSI Block PSI Combined PSI

M-F 6 4932 5069 5001F-M 6 5043 5034 5048F-F 6 5022 4952 4993M-M 14 4988 4975 4981

to settle down However in order to remove any remaining potential artifactu-al effects on EDA in Study 2 only visual instructions were given

Another difference in methodology concerned the screensaver viewed bythe participants during the experimental session In Study 1 some participantsreported that the series of patterns produced by the computer were being dis-played at too fast a rate such that some participants found it distracting or dis-turbing Again we felt that this may have an artifactual effect on EDA so wedecided for study two to reduce the speed of the screensaver

Thirdly in Study 2 we did not give participants detailed information abouttheir remote influence performance until the feedback letters that were distrib-uted at the conclusion of the entire study The reason for this change was mere-ly to streamline the procedure on what was already a time-consuming experi-ment Finally in another streamlining change we did not take VVIQ measuresin Study 2

Planned Hypotheses and Analyses

These were the same as with Study 1 That is hypothesis 1 predicted higherPSI scores for the co-operating condition than for the blocking condition(measured with a related t-test one-tailed) If no significant difference wasfound between blocking PSI and co-operating PSI scores the two would becombined to look at overall remote influence effect (hypothesis 2 single-meant-test one-tailed) If a significant difference was found with hypothesis 1 theanalysis for hypothesis 2 would be conducted with the data for co-operatingsessions only

Exploratory Questions

As with study one we took an exploratory variance measure of remote EDAinfluence We also gathered qualitative descriptions of the mental strategiesparticipants adopted whilst blocking and co-operating We also looked at sexeffects experimenter effects and post hoc influenceeinfluencer sex pairings

Participants

Study 2 had two experimenterinfluencers (Watt amp Ravenscroft) After aparticipant was recruited (usually by Ravenscroft) they were assigned to oneof the two experimenters for all three testing sessions This assignment wasdone unsystematically according to the preferences and availability of the ex-perimenters As planned 50 volunteer participants were drawn from the gener-al participant pool at the Koestler Parapsychology Unit and by word ofmouth All participants completed the Participant Information Form and aConsent form prior to the study The participants consisted of 33 females and17 males Average age was 41 years range from 22 years to 76 years of age Inorder to increase the studyrsquos relevance to the general population undergradu-ate psychology students were not recruited as participants The main criterion

Blocking vs Co-operating 527

528 C Watt et al

for participation was that the participants were interested in and open-mindedabout parapsychology Again there were no drop-outs from this study Allparticipants received a detailed personal feedback letter at the conclusion ofthe study

Study Two Results and Discussion

Remote Influence Measure

PSI scores are summarized in Table 6 which shows that the mean PSI scorewhen co-operating was nonsignificantly lower than mean PSI score whenblocking (t = - 595 df = 49) Hypothesis 1 was therefore not supported Asplanned in the absence of any significant difference between blocking and co-operating PSI scores the two conditions were combined to test hypothesis 2which predicted overall significant remote mental influence performance Thecombined PSI score was significantly greater than mean chance expectation of50 (t = 1806 df = 49 p = 038 1 - t) Therefore hypothesis 2 was supportedgiving evidence of an overall remote influence effect upon psychophysiologyin the prespecified direction

The failure to support hypothesis 1 confirms what we found in Study 1 thatthere was no significant difference between blocking and co-operating condi-tions on remote influence scores Indeed both studies found a slight effect inthe opposite direction to that predicted that is slightly higher PSI scores whenblocking was the influenceersquos intention than when co-operating was their in-tention It was encouraging to find significant support for a remote influenceeffect in the predicted direction for blocking and co-operating conditions com-bined No such trend was found in the previous study though with a smallernumber of participants (N = 32) the earlier study had lower statistical powerthan the present study

Exploratory Questions

Variance Measure

Table 7 shows the results for the exploratory measure of EDA variance lt 50represents reduced variance gt 50 represents increased variance We found nodifference between variance scores for cooperate versus block conditions(t = 150 df = 49) When these are combined the resulting figure does not de-viate from chance expectation (t = 440 df = 49) Thus as in Study 1 it seems

TABLE 6Summary Data Mean PSI Scores

(N = 50 MCE = 50 range in brackets)

Co-operate PSI Block PSI Combined PSI

5008 5019 5016(4821- 5221) (4807- 5557) (4900- 5296)

that the exploratory variance measure does not reveal any additional indica-tion of the operation of a remote mental influence in this study However thewider range of variance scoring continues to suggest that this measure mayhave greater potential sensitivity than the summary PSI score

Influenceesrsquo Blocking and Co-operating Strategies

We were interested in learning more about what particular strategies partici-pants chose to adopt when blocking and co-operating As in Study 1 a post-session questionnaire was used to gather this information Table 8 shows thepercentage of participants who adopted each kind of strategy

The table shows a very similar pattern to that found in Study 1 The mainpoint of interest is that almost no participants took a passive approach toblocking (ie adopting an openrelaxing strategy) A much higher proportionpreferred to adopt this approach when co-operating The preferred mode forboth conditions was to use some form of visual imagery such as imagining ashield or a screen when blocking and imagining an open channel such as ariver when co-operating

The majority of participants reported no concerns over the possibility ofhaving their physiology remotely influenced by another person as found instudy one

Sex Effects

As in study one the sexes were not equally represented among participantsWhereas Study 1 had an excess of male participants in Study 2 twice as manyfemales (N = 33) as males (N = 17) participated It was therefore important tolook for sex differences since any such differences could confound the inter-pretation of overall trends We found no difference between sexes when co-op-erating (female mean PSI score = 5008 male mean PSI score = 5006 t = 107 df = 48) There was a nonsignificant trend for females to have higher

Blocking vs Co-operating 529

TABLE 7Summary Data Mean Variance Scores(N = 50 MCE = 50 range in brackets)

Co-operate Variance Block Variance Combined Variance

5076 5054 5034(3164- 7279) (3572- 6813) (3749- 6671)

TABLE 8Influenceesrsquo Mental Strategies ()

Visual Imagery Cognitive VisCog OpenRelax

Blocking 64 26 6 2Co-operating 60 8 12 20

530 C Watt et al

PSI scores than males when blocking (mean PSI scores 5030 and 4998 re-spectively t = 992 df = 48) Combining blocking and co-operating conditionsthere was no overall difference between the sexes (mean PSI scores 5020 forfemales 5008 for males t = 665 df = 48)

Experimenter Effects

Study 2 found a statistically significant overall remote influence effect(combined PSI = 50155 t = 1806) Table 9 shows the breakdown of remoteinfluence scores for the two experimenters CW and JR Looking at the overalldata for the two experimenters we see that CWrsquos sessions were associatedwith most of the remote influence effect However due to the smaller N CWrsquossessions are not independently significant (t = 1632) and there is not a signifi-cant difference between CW and JRrsquos overall PSI scores (t = 678) Post hocwe calculated effect size r according to the formula r = [t2(t2+df)]12 for the twoexperimenters For JR (t = 817 df = 22) we found a small effect sizer = +17 For CW (t = 1632 df = 26) the effect size was moderate r = +30This may be some post hoc indication of an experimenter effect in Study 2

InfluenceeInfluencer Sex Pairings

As can be seen from Table 10 Study 2 found exactly the opposite trend tothat found in Study 1 On a post hoc test same sex pairings scored significant-ly higher on the remote influence PSI measure than mixed sex pairings(t = - 2034 df = 48 p = 048 2 - t) This contradicts Schlitz and LaBergersquos(1997) finding of greater autonomic detection of remote staring for mixed sexpairs

TABLE 9Remote Influence Results by Experimenter

Experimenter N sessions Coop PSI Block PSI Combined PSI

CW 27 5011 5026 5021JR 23 5003 5010 5009

TABLE 10Mean PSI Scores for Different InfluenceeInfluencer Sex Pairings

Pairing N Coop PSI Block PSI Combined PSI

MF 11 4977 5009 4997FM 17 4983 5021 5003FF 16 5035 5038 5038MM 6 5060 4978 5027

General Summary and Conclusions

We have described the findings of two studies exploring the limits of remotemental influence effects We wished to see whether the mental strategiesadopted by the influencee would affect remote influence scores In particularcould an influencee block or resist a remote influence attempt We recorded in-fluenceesrsquo electrodermal activity while the influencer attempted to calm to ac-tivate or to have no influence over the influenceesrsquo physiology In each studythere were two conditions reflecting the influenceesrsquo designated mental strate-gies in response to the remote influence attempt blocking and co-operating

There were only two formal hypotheses 1) that remote influence scoreswould be significantly higher during the co-operating condition than duringthe blocking condition and 2) that overall there would be significant evidenceof remote mental influence on physiology in the designated direction (iecalming when calming was the influencerrsquos intention activating when activat-ing was the influencerrsquos intention) Contrary to prediction Study 1 found thatremote influence effects were higher in the block condition than in the co-op-erate condition (but not to a significant extent) Also overall remote influenceperformance was at chance level This study therefore found no support for ei-ther of the formal hypotheses Study 2 consisted of a replication of Study 1with a few minor adjustments plus additional measures that were not reportedin this paper Again contrary to prediction remote influence scores were non-significantly higher in the block condition than in the co-operate conditionThis replicates the trend found in Study 1 failing to support hypothesis oneThese results suggest that the mental strategy adopted by the influencee in theface of a remote influence attempt has no effect on remote influence perfor-mance Referring back to the questions posed in the introduction then wehave not found any evidence to suggest that it is possible to resist a remote in-fluence effect and there is as yet no indication of the possible limiting condi-tions for remote mental influence

Unlike Study 1 Study 2 found statistically significant evidence of an overallremote influence effect in the prespecified direction (ie calming of EDAwhen that was the influencerrsquos intention activation of EDA when that was theinfluencerrsquos intention) In order to facilitate comparison with a recent meta-an-alytic review of remote influence studies we calculated the effect size (r) ac-cording to the formula r = [t2(t2+df)]12 For their review of 19 EDA influencestudies Schlitz and Braud (1997) found an average effect size of r = +25 OurStudy 1 with overall t = - 031 df = 31 gives a negligible effect size of r = - 006 Study 2 with overall t = 1806 df = 49 gives r = +25 Study 2 there-fore replicates the mean effect size found in EDA studies conducted by otherresearchers in other parapsychology laboratories Post hoc we can combinethe data of the two studies for greater statistical power Overall t = 1176 df = 81 giving a small effect size of r = +13

We can find no clear reason why Study 2 found an EDA influence effectwhile Study 1 did not A number of factors are worth considering 1) The

Blocking vs Co-operating 531

532 C Watt et al

significant result could be due to chance or undetected artifact 2) Study 2 im-plemented what we considered to be minor methodological improvements thatmight have removed some artifactual fluctuations from the EDA data that weobserved in study one 3) Each experimenter in Study 2 conducted three testingsessions with their participant By the third session in which the remote influ-ence measure was taken experimenters were on quite friendly terms with theirparticipants perhaps more so than in Study 1 where each participant came infor only one testing session 4) The participants in Study 2 were more represen-tative of the general population we felt than those in Study 1 Study 1 in-volved more undergraduate psychology students and associates of the experi-menters 5) The experimenters in study one were less experienced with theremote influence protocol than when they conducted Study 2

In addition to the formal hypotheses we reported a number of exploratorymeasures Study 1 found that influenceesrsquo self-reported vividness of visual im-agery appeared to be unrelated to the remote influence performance In bothstudies influenceesrsquo qualitative descriptions of the actual mental strategiesthey employed when blocking and co-operating revealed a tendency for differ-ent strategies to be preferred for the different conditions The use of visual im-agery was popular for both blocking (eg visualizing being surrounded by aprotective cocoon) and co-operating (eg visualizing a river flowing betweeninfluencer and influencee) However a considerable proportion of participantsreported adopting an open or relaxed frame of mind when co-operating whilstalmost no participants chose this strategy when blocking This was despite thefact that prior to each session we encouraged participants to adopt comparablestrategies for blocking and co-operating

It was interesting and perhaps surprising to note that in both studies themajority of participants reported no feelings of concern or discomfort at thepossibility of having their physiology remotely influenced Those who wereconcerned tended to be those who knew more about parapsychology and thepossible harmful applications of remote influence A small proportion of par-ticipants was unconcerned because they did not believe such an influence waspossible in the first place

Both studies employed an exploratory variance measure of EDA influenceThis measure showed a greater range of scoring overall thus perhaps greaterpotential sensitivity to a remote influence effect However the variance mea-sure showed no indication of a remote influence effect To date researchershave employed a variety of indicators of remote influence as the literaturegrows progress may be made on identifying which is the most appropriate in-dicator

Neither study found significant sex differences in remote influence effectsthough in both studies female influencees tended to have higher remote influ-ence scores than male influencees

In terms of their overall remote influence effects there appeared to be nodifference between experimenterinfluencers in Study 1 (r = +06 - 05 and

- 06) In Study 2 effect size measures indicated that the experimenters dif-fered in remote influence effect sizes (r = +17 and +30) Due to the smaller Nthis difference was not found to be statistically significant Post hoc combin-ing Studies 1 and 2 for experimenters JW and CW (ZMcD was not an experi-menter in Study 2) the effect sizes are r = +08 and r = +20 respectively

We conducted a post hoc analysis of the effects of influenceeinfluencer sexpairings on remote influence scores Our two studies came up with contradic-tory findings each statistically significant Study one found that mixed sexpairings had significantly higher remote influence scores than same sex pair-ings This replicated the post hoc finding of Schlitz and LaBerge (1997) in con-nection with the autonomic detection of remote observation Study 2 foundthat same sex pairings had significantly higher remote influence scores thanmixed sex pairings Our contradictory findings do not help to clarify the pic-ture on the importance or otherwise of sex pairings It is quite possible that sexis not the most salient aspect of the dynamic between pairs Age personalityand familiarity are likely to be important variables moderating the relationshipbetween influencer and influencee

There are many more variables that we could look at but as the number ofanalyses multiplies so does the opportunity for spurious significant resultsFor this reason we have named only two formal hypotheses for each experi-ment All other analyses are regarded as exploratory and any trends that wehave highlighted above would require replication

In conclusion these two studies have found no evidence of a difference inremote influence effects when influencees are asked to adopt cooperative men-tal strategies compared to when they are asked to adopt blocking mental strate-gies In fact contrary to expectation both studies found slightly higher remoteinfluence effects for the blocking condition compared to the co-operating con-dition On this basis we have found no indication of limiting conditions for re-mote influence effects Therefore while anecdotal advice exists on how to de-fend oneself against a remote physiological influence attempt in the realworld laboratory evidence to date has provided no support for this advice

Given the importance of this question both in practical terms and in theoret-ical terms we argue that this line of research deserves continued investigationIf a way could be found to conduct research with a more realistic scenario mdash where the remote influence was not just neutrally and abstractlydefined as unwelcome mdash then our research might have more meaningful thingsto say about the parameters of remote influence It is interesting and perhapssurprising to note that when Watt has lectured on this topic to informed lay au-diences the feedback has often been that researchers must use more potential-ly harmful or disruptive remote influence (of course with the informed consentof the participants) While there are clear ethical inhibitions when consideringpossible physiological influence perhaps research using higher level influencewould be more acceptable For example one could attempt to disrupt or en-hance the concentration of chess players who were aware of the influence

Blocking vs Co-operating 533

534 C Watt et al

attempt and who were concurrently blocking or co-operating with the influ-ence attempt Research already exists to suggest that remote influence does notjust operate at physiological levels Attention focusing for example can be fa-cilitated by a remote helper (Brady amp Morris 1997 Braud et al 1995) Futureresearch in this area must therefore expand beyond physiological influences

Acknowledgements

The authors gratefully acknowledge the financial support they receivedfrom the Institut fuumlr Grenzgebiete der Psychologie und PsychohygieneFreiburg Germany while conducting these studies We would also like tothank Dr Paul Stevens for generously helping to solve software difficultieswhen preparing for these studies and the anonymous reviewer for helpfulcomments Address reprint requests to Dr C Watt Department of PsychologyUniversity of Edinburgh 7 George Square Edinburgh EH8 9JZ Scotland UK

ReferencesBrady C amp Morris R L (1997) Attention focusing facilitated through remote mental interac-

tion A replication and exploration of parameters In R Wiseman (Ed) Proceedings of Pre-sented Papers The Parapsychological Association 40th Annual Convention held in conjunc-tion with The Society for Psychical Research pp 73ndash91

Braud W G (1985) Blockingshielding psychic functioning through psychological and psychictechniques A report of three preliminary studies In RA White and J Solfvin (Eds) Re-search in Parapsychology 1984 Metuchen NJ Scarecrow Press pp 42ndash44

Braud W G amp Schlitz M J (1991) Consciousness interactions with remote biological systemsAnomalous intentionality effects Subtle Energies 2 1

Braud W G Schlitz M J Collins J and Klitch H (1985) Further studies of the bio-PK effectFeedback blocking specificitygenerality In RA White amp J Solfvin (Eds) Research inParapsychology 1984 Metuchen NJ Scarecrow Press pp 45ndash48

Braud W G Shafer D McNeill K and Guerra V (1995) Attention focusing facilitatedthrough remote mental interaction Journal of the American Society for Psychical Research89 103

Cohen J (1997) Statistical Power Analysis for the Behavioral Sciences New York AcademicPress

Dalton K (1994) Remotely influenced ESP performance in a computer task A preliminarystudy Proceedings of the 37th Annual Convention of the Parapsychological Association pp95ndash103

Dossey L (1994) Healing and the mind Is there a dark side Journal of Scientific Exploration 873

Halifax-Grof J (1974) Hex death In A Angoff amp D Barth (Eds) Parapsychology and Anthro-pology Proceedings of an International Conference Held in London England August 29ndash311973 NY Parapsychology Foundation

Hope M (1983) Practical Techniques of Psychic Self-Defence Wellingborough Northampton-shire Aquarian Press

Marks DF (1973) Visual imagery differences in the recall of pictures British Journal of Psy-chology 64 17

Nash CB (1982) Psychokinetic control of bacterial growth Journal of the Society for PsychicalResearch 51 217

Radin DI Taylor RK and Braud WG (1995) Remote mental influence of human electroder-mal activity A pilot replication European Journal of Parapsychology 11 19

Schlitz M amp Braud W G (1997) Distant intentionality and healing Assessing the evidence Al-ternative Therapies 3 62

Schlitz M J amp LaBerge S (1997) Covert observation increases skin conductance in subjectsunaware of when they are being observed A replication Journal of Parapsychology 61 185

Schmeidler G R (1958) Agent-percipient relationships Journal of the American Society forPsychical Research 52 47

Schmeidler G R (1961) Evidence for two kinds of telepathy International Journal of Parapsy-chology 3 5

Tart C T (1979) A survey of expert opinion on potentially negative uses of psi United Statesgovernment interest in psi and the level of research funding of the field In WG Roll (Ed) Re-search in Parapsychology 1978 Metuchen NJ Scarecrow Press pp 54

Tart C T amp Labore C M (1986) Attitudes towards strongly functioning psi A preliminary sur-vey Journal of the America Society for Psychical Research 80 163

Tart C T (1986) Psychicsrsquo fears of psychic powers Journal of the American Society for Psychi-cal Research 80 278

Watt C A McDermott Z amp Ravenscroft J (1997) Exploring the limits of remote mental inter-action A comparison of blocking versus co-operating strategies In R Wiseman (Ed) Pro-ceedings of the 40th Annual Convention of the Parapsychological Association held in conjunc-tion with the Society for Psychical Research Brighton England pp 464ndash476

Blocking vs Co-operating 535

imagers) to use visual imagery strategies Those who had high VVIQ scoresalso used visual imagery but more often used this in combination with othertactics

Other Qualitative Reports from Participants According to their post-ses-sion questionnaire responses participants generally felt comfortable about theattempt to be influenced (N = 24) Three participants said they felt apprehen-sive or uncomfortable with the idea of being influenced Interestingly of thesethree two were actively involved with parapsychology and were aware of theharmful psi implications discussed at the beginning of this report The thirdwas a psychology student who had attended a lecture given by Watt in whichthe possibilities of harmful psi were discussed Another group of three partici-pants said they would be uncomfortable with the idea of being influenced ifthey believed such influence was possible

Seven participants said they found it was difficult to sustain concentrationfor the 10 minutes that comprised each condition Five participants found thescreensaver distracting or disturbing Two participants specifically mentionedthat they found blocking a more taxing task than co-operating

Sex Effects

There were more males (N = 20) than females (N = 12) in this study howev-er there was no significant difference in PSI scores for remote physiology in-fluence between males and females (male PSI vs female PSI t = 1085 male PSI = 4987 female PSI = 5021)

Experimenter Effects

We studied whether there was any difference in PSI scores obtained by thethree experimenters The results are shown in Table 4 There seems to be nosign of experimenter effects in study one

InfluenceeInfluencer Sex Pairings

Table 5 gives the mean PSI scores for different influenceeinfluencer pair-ings Taking the combined PSI scores it can be seen that the highest remote in-fluence effect is for FM pairs followed by MF pairs followed by FF pairsand the lowest scoring is for MM pairs Post hoc it was found that the mixedsex pairings showed significantly higher remote influence scores than the same

Blocking vs Co-operating 525

TABLE 4Remote Influence Results by Experimenter

Experimenter N sessions Coop PSI Block PSI Combined PSI

CW 12 4983 5011 4997ZMcD 10 5015 4998 5006JR 10 4994 4999 4996

526 C Watt et al

sex pairings (t = - 2016 df = 30 p = 053 2 - t) This replicates the effect foundby Schlitz and LaBerge (1997)

Study One Conclusions

In Study 1 we developed a methodology to examine intentional resistance toa remote mental influence attempt (Watt et al 1997) We contrasted influ-enceesrsquo physiological activity when they were mentally co-operating with aremote attempt to influence their physiology with physiological activity wheninfluencees were mentally blocking the influence attempt We found no evi-dence for any remote influence effect overall and no difference betweenblocking and co-operating conditions Our qualitative measures yielded someinteresting information about participantsrsquo mental strategies and reactions dur-ing the remote influence session We found no difference between experi-menters in remote influence scores but we did find that mixed sex influ-enceeinfluencer pairings scored more highly than same sex pairings

Study Two Method

Study 2 was planned to explore in more depth the nature of the relationshipbetween aspects of personality and cognition and psi performance Partici-pants took part in three testing sessions approximately one week apart Thefirst consisted of various psychological measures the second consisted of aforced choice ESP task and the third consisted of a replication of the blocking-co-operating study (Study 1) described above The psychological measureswere not scored until after completion of the remote influence session Partici-pants received feedback about their forced choice ESP performance at the con-clusion of the second session For ease of interpretation the present paper isrestricted to a description of the third testing session which was essentially areplication of study one with fewer influencers and greater statistical power

We introduced some slight changes to the methodology of Study 2 based onour experiences with Study 1 At the start of each of the two runs in Study 1the participant was given auditory instructions either to ldquoblock nowrdquo or ldquoco-operate nowrdquo We observed that the auditory instructions appeared to affect theparticipantrsquos EDA as if they were startled As planned the first recordingepoch in each run was not used in analysis to give the participantrsquos EDA time

TABLE 5Mean PSI Scores for Different InfluenceeInfluencer Sex Pairings

Pairing N Coop PSI Block PSI Combined PSI

M-F 6 4932 5069 5001F-M 6 5043 5034 5048F-F 6 5022 4952 4993M-M 14 4988 4975 4981

to settle down However in order to remove any remaining potential artifactu-al effects on EDA in Study 2 only visual instructions were given

Another difference in methodology concerned the screensaver viewed bythe participants during the experimental session In Study 1 some participantsreported that the series of patterns produced by the computer were being dis-played at too fast a rate such that some participants found it distracting or dis-turbing Again we felt that this may have an artifactual effect on EDA so wedecided for study two to reduce the speed of the screensaver

Thirdly in Study 2 we did not give participants detailed information abouttheir remote influence performance until the feedback letters that were distrib-uted at the conclusion of the entire study The reason for this change was mere-ly to streamline the procedure on what was already a time-consuming experi-ment Finally in another streamlining change we did not take VVIQ measuresin Study 2

Planned Hypotheses and Analyses

These were the same as with Study 1 That is hypothesis 1 predicted higherPSI scores for the co-operating condition than for the blocking condition(measured with a related t-test one-tailed) If no significant difference wasfound between blocking PSI and co-operating PSI scores the two would becombined to look at overall remote influence effect (hypothesis 2 single-meant-test one-tailed) If a significant difference was found with hypothesis 1 theanalysis for hypothesis 2 would be conducted with the data for co-operatingsessions only

Exploratory Questions

As with study one we took an exploratory variance measure of remote EDAinfluence We also gathered qualitative descriptions of the mental strategiesparticipants adopted whilst blocking and co-operating We also looked at sexeffects experimenter effects and post hoc influenceeinfluencer sex pairings

Participants

Study 2 had two experimenterinfluencers (Watt amp Ravenscroft) After aparticipant was recruited (usually by Ravenscroft) they were assigned to oneof the two experimenters for all three testing sessions This assignment wasdone unsystematically according to the preferences and availability of the ex-perimenters As planned 50 volunteer participants were drawn from the gener-al participant pool at the Koestler Parapsychology Unit and by word ofmouth All participants completed the Participant Information Form and aConsent form prior to the study The participants consisted of 33 females and17 males Average age was 41 years range from 22 years to 76 years of age Inorder to increase the studyrsquos relevance to the general population undergradu-ate psychology students were not recruited as participants The main criterion

Blocking vs Co-operating 527

528 C Watt et al

for participation was that the participants were interested in and open-mindedabout parapsychology Again there were no drop-outs from this study Allparticipants received a detailed personal feedback letter at the conclusion ofthe study

Study Two Results and Discussion

Remote Influence Measure

PSI scores are summarized in Table 6 which shows that the mean PSI scorewhen co-operating was nonsignificantly lower than mean PSI score whenblocking (t = - 595 df = 49) Hypothesis 1 was therefore not supported Asplanned in the absence of any significant difference between blocking and co-operating PSI scores the two conditions were combined to test hypothesis 2which predicted overall significant remote mental influence performance Thecombined PSI score was significantly greater than mean chance expectation of50 (t = 1806 df = 49 p = 038 1 - t) Therefore hypothesis 2 was supportedgiving evidence of an overall remote influence effect upon psychophysiologyin the prespecified direction

The failure to support hypothesis 1 confirms what we found in Study 1 thatthere was no significant difference between blocking and co-operating condi-tions on remote influence scores Indeed both studies found a slight effect inthe opposite direction to that predicted that is slightly higher PSI scores whenblocking was the influenceersquos intention than when co-operating was their in-tention It was encouraging to find significant support for a remote influenceeffect in the predicted direction for blocking and co-operating conditions com-bined No such trend was found in the previous study though with a smallernumber of participants (N = 32) the earlier study had lower statistical powerthan the present study

Exploratory Questions

Variance Measure

Table 7 shows the results for the exploratory measure of EDA variance lt 50represents reduced variance gt 50 represents increased variance We found nodifference between variance scores for cooperate versus block conditions(t = 150 df = 49) When these are combined the resulting figure does not de-viate from chance expectation (t = 440 df = 49) Thus as in Study 1 it seems

TABLE 6Summary Data Mean PSI Scores

(N = 50 MCE = 50 range in brackets)

Co-operate PSI Block PSI Combined PSI

5008 5019 5016(4821- 5221) (4807- 5557) (4900- 5296)

that the exploratory variance measure does not reveal any additional indica-tion of the operation of a remote mental influence in this study However thewider range of variance scoring continues to suggest that this measure mayhave greater potential sensitivity than the summary PSI score

Influenceesrsquo Blocking and Co-operating Strategies

We were interested in learning more about what particular strategies partici-pants chose to adopt when blocking and co-operating As in Study 1 a post-session questionnaire was used to gather this information Table 8 shows thepercentage of participants who adopted each kind of strategy

The table shows a very similar pattern to that found in Study 1 The mainpoint of interest is that almost no participants took a passive approach toblocking (ie adopting an openrelaxing strategy) A much higher proportionpreferred to adopt this approach when co-operating The preferred mode forboth conditions was to use some form of visual imagery such as imagining ashield or a screen when blocking and imagining an open channel such as ariver when co-operating

The majority of participants reported no concerns over the possibility ofhaving their physiology remotely influenced by another person as found instudy one

Sex Effects

As in study one the sexes were not equally represented among participantsWhereas Study 1 had an excess of male participants in Study 2 twice as manyfemales (N = 33) as males (N = 17) participated It was therefore important tolook for sex differences since any such differences could confound the inter-pretation of overall trends We found no difference between sexes when co-op-erating (female mean PSI score = 5008 male mean PSI score = 5006 t = 107 df = 48) There was a nonsignificant trend for females to have higher

Blocking vs Co-operating 529

TABLE 7Summary Data Mean Variance Scores(N = 50 MCE = 50 range in brackets)

Co-operate Variance Block Variance Combined Variance

5076 5054 5034(3164- 7279) (3572- 6813) (3749- 6671)

TABLE 8Influenceesrsquo Mental Strategies ()

Visual Imagery Cognitive VisCog OpenRelax

Blocking 64 26 6 2Co-operating 60 8 12 20

530 C Watt et al

PSI scores than males when blocking (mean PSI scores 5030 and 4998 re-spectively t = 992 df = 48) Combining blocking and co-operating conditionsthere was no overall difference between the sexes (mean PSI scores 5020 forfemales 5008 for males t = 665 df = 48)

Experimenter Effects

Study 2 found a statistically significant overall remote influence effect(combined PSI = 50155 t = 1806) Table 9 shows the breakdown of remoteinfluence scores for the two experimenters CW and JR Looking at the overalldata for the two experimenters we see that CWrsquos sessions were associatedwith most of the remote influence effect However due to the smaller N CWrsquossessions are not independently significant (t = 1632) and there is not a signifi-cant difference between CW and JRrsquos overall PSI scores (t = 678) Post hocwe calculated effect size r according to the formula r = [t2(t2+df)]12 for the twoexperimenters For JR (t = 817 df = 22) we found a small effect sizer = +17 For CW (t = 1632 df = 26) the effect size was moderate r = +30This may be some post hoc indication of an experimenter effect in Study 2

InfluenceeInfluencer Sex Pairings

As can be seen from Table 10 Study 2 found exactly the opposite trend tothat found in Study 1 On a post hoc test same sex pairings scored significant-ly higher on the remote influence PSI measure than mixed sex pairings(t = - 2034 df = 48 p = 048 2 - t) This contradicts Schlitz and LaBergersquos(1997) finding of greater autonomic detection of remote staring for mixed sexpairs

TABLE 9Remote Influence Results by Experimenter

Experimenter N sessions Coop PSI Block PSI Combined PSI

CW 27 5011 5026 5021JR 23 5003 5010 5009

TABLE 10Mean PSI Scores for Different InfluenceeInfluencer Sex Pairings

Pairing N Coop PSI Block PSI Combined PSI

MF 11 4977 5009 4997FM 17 4983 5021 5003FF 16 5035 5038 5038MM 6 5060 4978 5027

General Summary and Conclusions

We have described the findings of two studies exploring the limits of remotemental influence effects We wished to see whether the mental strategiesadopted by the influencee would affect remote influence scores In particularcould an influencee block or resist a remote influence attempt We recorded in-fluenceesrsquo electrodermal activity while the influencer attempted to calm to ac-tivate or to have no influence over the influenceesrsquo physiology In each studythere were two conditions reflecting the influenceesrsquo designated mental strate-gies in response to the remote influence attempt blocking and co-operating

There were only two formal hypotheses 1) that remote influence scoreswould be significantly higher during the co-operating condition than duringthe blocking condition and 2) that overall there would be significant evidenceof remote mental influence on physiology in the designated direction (iecalming when calming was the influencerrsquos intention activating when activat-ing was the influencerrsquos intention) Contrary to prediction Study 1 found thatremote influence effects were higher in the block condition than in the co-op-erate condition (but not to a significant extent) Also overall remote influenceperformance was at chance level This study therefore found no support for ei-ther of the formal hypotheses Study 2 consisted of a replication of Study 1with a few minor adjustments plus additional measures that were not reportedin this paper Again contrary to prediction remote influence scores were non-significantly higher in the block condition than in the co-operate conditionThis replicates the trend found in Study 1 failing to support hypothesis oneThese results suggest that the mental strategy adopted by the influencee in theface of a remote influence attempt has no effect on remote influence perfor-mance Referring back to the questions posed in the introduction then wehave not found any evidence to suggest that it is possible to resist a remote in-fluence effect and there is as yet no indication of the possible limiting condi-tions for remote mental influence

Unlike Study 1 Study 2 found statistically significant evidence of an overallremote influence effect in the prespecified direction (ie calming of EDAwhen that was the influencerrsquos intention activation of EDA when that was theinfluencerrsquos intention) In order to facilitate comparison with a recent meta-an-alytic review of remote influence studies we calculated the effect size (r) ac-cording to the formula r = [t2(t2+df)]12 For their review of 19 EDA influencestudies Schlitz and Braud (1997) found an average effect size of r = +25 OurStudy 1 with overall t = - 031 df = 31 gives a negligible effect size of r = - 006 Study 2 with overall t = 1806 df = 49 gives r = +25 Study 2 there-fore replicates the mean effect size found in EDA studies conducted by otherresearchers in other parapsychology laboratories Post hoc we can combinethe data of the two studies for greater statistical power Overall t = 1176 df = 81 giving a small effect size of r = +13

We can find no clear reason why Study 2 found an EDA influence effectwhile Study 1 did not A number of factors are worth considering 1) The

Blocking vs Co-operating 531

532 C Watt et al

significant result could be due to chance or undetected artifact 2) Study 2 im-plemented what we considered to be minor methodological improvements thatmight have removed some artifactual fluctuations from the EDA data that weobserved in study one 3) Each experimenter in Study 2 conducted three testingsessions with their participant By the third session in which the remote influ-ence measure was taken experimenters were on quite friendly terms with theirparticipants perhaps more so than in Study 1 where each participant came infor only one testing session 4) The participants in Study 2 were more represen-tative of the general population we felt than those in Study 1 Study 1 in-volved more undergraduate psychology students and associates of the experi-menters 5) The experimenters in study one were less experienced with theremote influence protocol than when they conducted Study 2

In addition to the formal hypotheses we reported a number of exploratorymeasures Study 1 found that influenceesrsquo self-reported vividness of visual im-agery appeared to be unrelated to the remote influence performance In bothstudies influenceesrsquo qualitative descriptions of the actual mental strategiesthey employed when blocking and co-operating revealed a tendency for differ-ent strategies to be preferred for the different conditions The use of visual im-agery was popular for both blocking (eg visualizing being surrounded by aprotective cocoon) and co-operating (eg visualizing a river flowing betweeninfluencer and influencee) However a considerable proportion of participantsreported adopting an open or relaxed frame of mind when co-operating whilstalmost no participants chose this strategy when blocking This was despite thefact that prior to each session we encouraged participants to adopt comparablestrategies for blocking and co-operating

It was interesting and perhaps surprising to note that in both studies themajority of participants reported no feelings of concern or discomfort at thepossibility of having their physiology remotely influenced Those who wereconcerned tended to be those who knew more about parapsychology and thepossible harmful applications of remote influence A small proportion of par-ticipants was unconcerned because they did not believe such an influence waspossible in the first place

Both studies employed an exploratory variance measure of EDA influenceThis measure showed a greater range of scoring overall thus perhaps greaterpotential sensitivity to a remote influence effect However the variance mea-sure showed no indication of a remote influence effect To date researchershave employed a variety of indicators of remote influence as the literaturegrows progress may be made on identifying which is the most appropriate in-dicator

Neither study found significant sex differences in remote influence effectsthough in both studies female influencees tended to have higher remote influ-ence scores than male influencees

In terms of their overall remote influence effects there appeared to be nodifference between experimenterinfluencers in Study 1 (r = +06 - 05 and

- 06) In Study 2 effect size measures indicated that the experimenters dif-fered in remote influence effect sizes (r = +17 and +30) Due to the smaller Nthis difference was not found to be statistically significant Post hoc combin-ing Studies 1 and 2 for experimenters JW and CW (ZMcD was not an experi-menter in Study 2) the effect sizes are r = +08 and r = +20 respectively

We conducted a post hoc analysis of the effects of influenceeinfluencer sexpairings on remote influence scores Our two studies came up with contradic-tory findings each statistically significant Study one found that mixed sexpairings had significantly higher remote influence scores than same sex pair-ings This replicated the post hoc finding of Schlitz and LaBerge (1997) in con-nection with the autonomic detection of remote observation Study 2 foundthat same sex pairings had significantly higher remote influence scores thanmixed sex pairings Our contradictory findings do not help to clarify the pic-ture on the importance or otherwise of sex pairings It is quite possible that sexis not the most salient aspect of the dynamic between pairs Age personalityand familiarity are likely to be important variables moderating the relationshipbetween influencer and influencee

There are many more variables that we could look at but as the number ofanalyses multiplies so does the opportunity for spurious significant resultsFor this reason we have named only two formal hypotheses for each experi-ment All other analyses are regarded as exploratory and any trends that wehave highlighted above would require replication

In conclusion these two studies have found no evidence of a difference inremote influence effects when influencees are asked to adopt cooperative men-tal strategies compared to when they are asked to adopt blocking mental strate-gies In fact contrary to expectation both studies found slightly higher remoteinfluence effects for the blocking condition compared to the co-operating con-dition On this basis we have found no indication of limiting conditions for re-mote influence effects Therefore while anecdotal advice exists on how to de-fend oneself against a remote physiological influence attempt in the realworld laboratory evidence to date has provided no support for this advice

Given the importance of this question both in practical terms and in theoret-ical terms we argue that this line of research deserves continued investigationIf a way could be found to conduct research with a more realistic scenario mdash where the remote influence was not just neutrally and abstractlydefined as unwelcome mdash then our research might have more meaningful thingsto say about the parameters of remote influence It is interesting and perhapssurprising to note that when Watt has lectured on this topic to informed lay au-diences the feedback has often been that researchers must use more potential-ly harmful or disruptive remote influence (of course with the informed consentof the participants) While there are clear ethical inhibitions when consideringpossible physiological influence perhaps research using higher level influencewould be more acceptable For example one could attempt to disrupt or en-hance the concentration of chess players who were aware of the influence

Blocking vs Co-operating 533

534 C Watt et al

attempt and who were concurrently blocking or co-operating with the influ-ence attempt Research already exists to suggest that remote influence does notjust operate at physiological levels Attention focusing for example can be fa-cilitated by a remote helper (Brady amp Morris 1997 Braud et al 1995) Futureresearch in this area must therefore expand beyond physiological influences

Acknowledgements

The authors gratefully acknowledge the financial support they receivedfrom the Institut fuumlr Grenzgebiete der Psychologie und PsychohygieneFreiburg Germany while conducting these studies We would also like tothank Dr Paul Stevens for generously helping to solve software difficultieswhen preparing for these studies and the anonymous reviewer for helpfulcomments Address reprint requests to Dr C Watt Department of PsychologyUniversity of Edinburgh 7 George Square Edinburgh EH8 9JZ Scotland UK

ReferencesBrady C amp Morris R L (1997) Attention focusing facilitated through remote mental interac-

tion A replication and exploration of parameters In R Wiseman (Ed) Proceedings of Pre-sented Papers The Parapsychological Association 40th Annual Convention held in conjunc-tion with The Society for Psychical Research pp 73ndash91

Braud W G (1985) Blockingshielding psychic functioning through psychological and psychictechniques A report of three preliminary studies In RA White and J Solfvin (Eds) Re-search in Parapsychology 1984 Metuchen NJ Scarecrow Press pp 42ndash44

Braud W G amp Schlitz M J (1991) Consciousness interactions with remote biological systemsAnomalous intentionality effects Subtle Energies 2 1

Braud W G Schlitz M J Collins J and Klitch H (1985) Further studies of the bio-PK effectFeedback blocking specificitygenerality In RA White amp J Solfvin (Eds) Research inParapsychology 1984 Metuchen NJ Scarecrow Press pp 45ndash48

Braud W G Shafer D McNeill K and Guerra V (1995) Attention focusing facilitatedthrough remote mental interaction Journal of the American Society for Psychical Research89 103

Cohen J (1997) Statistical Power Analysis for the Behavioral Sciences New York AcademicPress

Dalton K (1994) Remotely influenced ESP performance in a computer task A preliminarystudy Proceedings of the 37th Annual Convention of the Parapsychological Association pp95ndash103

Dossey L (1994) Healing and the mind Is there a dark side Journal of Scientific Exploration 873

Halifax-Grof J (1974) Hex death In A Angoff amp D Barth (Eds) Parapsychology and Anthro-pology Proceedings of an International Conference Held in London England August 29ndash311973 NY Parapsychology Foundation

Hope M (1983) Practical Techniques of Psychic Self-Defence Wellingborough Northampton-shire Aquarian Press

Marks DF (1973) Visual imagery differences in the recall of pictures British Journal of Psy-chology 64 17

Nash CB (1982) Psychokinetic control of bacterial growth Journal of the Society for PsychicalResearch 51 217

Radin DI Taylor RK and Braud WG (1995) Remote mental influence of human electroder-mal activity A pilot replication European Journal of Parapsychology 11 19

Schlitz M amp Braud W G (1997) Distant intentionality and healing Assessing the evidence Al-ternative Therapies 3 62

Schlitz M J amp LaBerge S (1997) Covert observation increases skin conductance in subjectsunaware of when they are being observed A replication Journal of Parapsychology 61 185

Schmeidler G R (1958) Agent-percipient relationships Journal of the American Society forPsychical Research 52 47

Schmeidler G R (1961) Evidence for two kinds of telepathy International Journal of Parapsy-chology 3 5

Tart C T (1979) A survey of expert opinion on potentially negative uses of psi United Statesgovernment interest in psi and the level of research funding of the field In WG Roll (Ed) Re-search in Parapsychology 1978 Metuchen NJ Scarecrow Press pp 54

Tart C T amp Labore C M (1986) Attitudes towards strongly functioning psi A preliminary sur-vey Journal of the America Society for Psychical Research 80 163

Tart C T (1986) Psychicsrsquo fears of psychic powers Journal of the American Society for Psychi-cal Research 80 278

Watt C A McDermott Z amp Ravenscroft J (1997) Exploring the limits of remote mental inter-action A comparison of blocking versus co-operating strategies In R Wiseman (Ed) Pro-ceedings of the 40th Annual Convention of the Parapsychological Association held in conjunc-tion with the Society for Psychical Research Brighton England pp 464ndash476

Blocking vs Co-operating 535

526 C Watt et al

sex pairings (t = - 2016 df = 30 p = 053 2 - t) This replicates the effect foundby Schlitz and LaBerge (1997)

Study One Conclusions

In Study 1 we developed a methodology to examine intentional resistance toa remote mental influence attempt (Watt et al 1997) We contrasted influ-enceesrsquo physiological activity when they were mentally co-operating with aremote attempt to influence their physiology with physiological activity wheninfluencees were mentally blocking the influence attempt We found no evi-dence for any remote influence effect overall and no difference betweenblocking and co-operating conditions Our qualitative measures yielded someinteresting information about participantsrsquo mental strategies and reactions dur-ing the remote influence session We found no difference between experi-menters in remote influence scores but we did find that mixed sex influ-enceeinfluencer pairings scored more highly than same sex pairings

Study Two Method

Study 2 was planned to explore in more depth the nature of the relationshipbetween aspects of personality and cognition and psi performance Partici-pants took part in three testing sessions approximately one week apart Thefirst consisted of various psychological measures the second consisted of aforced choice ESP task and the third consisted of a replication of the blocking-co-operating study (Study 1) described above The psychological measureswere not scored until after completion of the remote influence session Partici-pants received feedback about their forced choice ESP performance at the con-clusion of the second session For ease of interpretation the present paper isrestricted to a description of the third testing session which was essentially areplication of study one with fewer influencers and greater statistical power

We introduced some slight changes to the methodology of Study 2 based onour experiences with Study 1 At the start of each of the two runs in Study 1the participant was given auditory instructions either to ldquoblock nowrdquo or ldquoco-operate nowrdquo We observed that the auditory instructions appeared to affect theparticipantrsquos EDA as if they were startled As planned the first recordingepoch in each run was not used in analysis to give the participantrsquos EDA time

TABLE 5Mean PSI Scores for Different InfluenceeInfluencer Sex Pairings

Pairing N Coop PSI Block PSI Combined PSI

M-F 6 4932 5069 5001F-M 6 5043 5034 5048F-F 6 5022 4952 4993M-M 14 4988 4975 4981

to settle down However in order to remove any remaining potential artifactu-al effects on EDA in Study 2 only visual instructions were given

Another difference in methodology concerned the screensaver viewed bythe participants during the experimental session In Study 1 some participantsreported that the series of patterns produced by the computer were being dis-played at too fast a rate such that some participants found it distracting or dis-turbing Again we felt that this may have an artifactual effect on EDA so wedecided for study two to reduce the speed of the screensaver

Thirdly in Study 2 we did not give participants detailed information abouttheir remote influence performance until the feedback letters that were distrib-uted at the conclusion of the entire study The reason for this change was mere-ly to streamline the procedure on what was already a time-consuming experi-ment Finally in another streamlining change we did not take VVIQ measuresin Study 2

Planned Hypotheses and Analyses

These were the same as with Study 1 That is hypothesis 1 predicted higherPSI scores for the co-operating condition than for the blocking condition(measured with a related t-test one-tailed) If no significant difference wasfound between blocking PSI and co-operating PSI scores the two would becombined to look at overall remote influence effect (hypothesis 2 single-meant-test one-tailed) If a significant difference was found with hypothesis 1 theanalysis for hypothesis 2 would be conducted with the data for co-operatingsessions only

Exploratory Questions

As with study one we took an exploratory variance measure of remote EDAinfluence We also gathered qualitative descriptions of the mental strategiesparticipants adopted whilst blocking and co-operating We also looked at sexeffects experimenter effects and post hoc influenceeinfluencer sex pairings

Participants

Study 2 had two experimenterinfluencers (Watt amp Ravenscroft) After aparticipant was recruited (usually by Ravenscroft) they were assigned to oneof the two experimenters for all three testing sessions This assignment wasdone unsystematically according to the preferences and availability of the ex-perimenters As planned 50 volunteer participants were drawn from the gener-al participant pool at the Koestler Parapsychology Unit and by word ofmouth All participants completed the Participant Information Form and aConsent form prior to the study The participants consisted of 33 females and17 males Average age was 41 years range from 22 years to 76 years of age Inorder to increase the studyrsquos relevance to the general population undergradu-ate psychology students were not recruited as participants The main criterion

Blocking vs Co-operating 527

528 C Watt et al

for participation was that the participants were interested in and open-mindedabout parapsychology Again there were no drop-outs from this study Allparticipants received a detailed personal feedback letter at the conclusion ofthe study

Study Two Results and Discussion

Remote Influence Measure

PSI scores are summarized in Table 6 which shows that the mean PSI scorewhen co-operating was nonsignificantly lower than mean PSI score whenblocking (t = - 595 df = 49) Hypothesis 1 was therefore not supported Asplanned in the absence of any significant difference between blocking and co-operating PSI scores the two conditions were combined to test hypothesis 2which predicted overall significant remote mental influence performance Thecombined PSI score was significantly greater than mean chance expectation of50 (t = 1806 df = 49 p = 038 1 - t) Therefore hypothesis 2 was supportedgiving evidence of an overall remote influence effect upon psychophysiologyin the prespecified direction

The failure to support hypothesis 1 confirms what we found in Study 1 thatthere was no significant difference between blocking and co-operating condi-tions on remote influence scores Indeed both studies found a slight effect inthe opposite direction to that predicted that is slightly higher PSI scores whenblocking was the influenceersquos intention than when co-operating was their in-tention It was encouraging to find significant support for a remote influenceeffect in the predicted direction for blocking and co-operating conditions com-bined No such trend was found in the previous study though with a smallernumber of participants (N = 32) the earlier study had lower statistical powerthan the present study

Exploratory Questions

Variance Measure

Table 7 shows the results for the exploratory measure of EDA variance lt 50represents reduced variance gt 50 represents increased variance We found nodifference between variance scores for cooperate versus block conditions(t = 150 df = 49) When these are combined the resulting figure does not de-viate from chance expectation (t = 440 df = 49) Thus as in Study 1 it seems

TABLE 6Summary Data Mean PSI Scores

(N = 50 MCE = 50 range in brackets)

Co-operate PSI Block PSI Combined PSI

5008 5019 5016(4821- 5221) (4807- 5557) (4900- 5296)

that the exploratory variance measure does not reveal any additional indica-tion of the operation of a remote mental influence in this study However thewider range of variance scoring continues to suggest that this measure mayhave greater potential sensitivity than the summary PSI score

Influenceesrsquo Blocking and Co-operating Strategies

We were interested in learning more about what particular strategies partici-pants chose to adopt when blocking and co-operating As in Study 1 a post-session questionnaire was used to gather this information Table 8 shows thepercentage of participants who adopted each kind of strategy

The table shows a very similar pattern to that found in Study 1 The mainpoint of interest is that almost no participants took a passive approach toblocking (ie adopting an openrelaxing strategy) A much higher proportionpreferred to adopt this approach when co-operating The preferred mode forboth conditions was to use some form of visual imagery such as imagining ashield or a screen when blocking and imagining an open channel such as ariver when co-operating

The majority of participants reported no concerns over the possibility ofhaving their physiology remotely influenced by another person as found instudy one

Sex Effects

As in study one the sexes were not equally represented among participantsWhereas Study 1 had an excess of male participants in Study 2 twice as manyfemales (N = 33) as males (N = 17) participated It was therefore important tolook for sex differences since any such differences could confound the inter-pretation of overall trends We found no difference between sexes when co-op-erating (female mean PSI score = 5008 male mean PSI score = 5006 t = 107 df = 48) There was a nonsignificant trend for females to have higher

Blocking vs Co-operating 529

TABLE 7Summary Data Mean Variance Scores(N = 50 MCE = 50 range in brackets)

Co-operate Variance Block Variance Combined Variance

5076 5054 5034(3164- 7279) (3572- 6813) (3749- 6671)

TABLE 8Influenceesrsquo Mental Strategies ()

Visual Imagery Cognitive VisCog OpenRelax

Blocking 64 26 6 2Co-operating 60 8 12 20

530 C Watt et al

PSI scores than males when blocking (mean PSI scores 5030 and 4998 re-spectively t = 992 df = 48) Combining blocking and co-operating conditionsthere was no overall difference between the sexes (mean PSI scores 5020 forfemales 5008 for males t = 665 df = 48)

Experimenter Effects

Study 2 found a statistically significant overall remote influence effect(combined PSI = 50155 t = 1806) Table 9 shows the breakdown of remoteinfluence scores for the two experimenters CW and JR Looking at the overalldata for the two experimenters we see that CWrsquos sessions were associatedwith most of the remote influence effect However due to the smaller N CWrsquossessions are not independently significant (t = 1632) and there is not a signifi-cant difference between CW and JRrsquos overall PSI scores (t = 678) Post hocwe calculated effect size r according to the formula r = [t2(t2+df)]12 for the twoexperimenters For JR (t = 817 df = 22) we found a small effect sizer = +17 For CW (t = 1632 df = 26) the effect size was moderate r = +30This may be some post hoc indication of an experimenter effect in Study 2

InfluenceeInfluencer Sex Pairings

As can be seen from Table 10 Study 2 found exactly the opposite trend tothat found in Study 1 On a post hoc test same sex pairings scored significant-ly higher on the remote influence PSI measure than mixed sex pairings(t = - 2034 df = 48 p = 048 2 - t) This contradicts Schlitz and LaBergersquos(1997) finding of greater autonomic detection of remote staring for mixed sexpairs

TABLE 9Remote Influence Results by Experimenter

Experimenter N sessions Coop PSI Block PSI Combined PSI

CW 27 5011 5026 5021JR 23 5003 5010 5009

TABLE 10Mean PSI Scores for Different InfluenceeInfluencer Sex Pairings

Pairing N Coop PSI Block PSI Combined PSI

MF 11 4977 5009 4997FM 17 4983 5021 5003FF 16 5035 5038 5038MM 6 5060 4978 5027

General Summary and Conclusions

We have described the findings of two studies exploring the limits of remotemental influence effects We wished to see whether the mental strategiesadopted by the influencee would affect remote influence scores In particularcould an influencee block or resist a remote influence attempt We recorded in-fluenceesrsquo electrodermal activity while the influencer attempted to calm to ac-tivate or to have no influence over the influenceesrsquo physiology In each studythere were two conditions reflecting the influenceesrsquo designated mental strate-gies in response to the remote influence attempt blocking and co-operating

There were only two formal hypotheses 1) that remote influence scoreswould be significantly higher during the co-operating condition than duringthe blocking condition and 2) that overall there would be significant evidenceof remote mental influence on physiology in the designated direction (iecalming when calming was the influencerrsquos intention activating when activat-ing was the influencerrsquos intention) Contrary to prediction Study 1 found thatremote influence effects were higher in the block condition than in the co-op-erate condition (but not to a significant extent) Also overall remote influenceperformance was at chance level This study therefore found no support for ei-ther of the formal hypotheses Study 2 consisted of a replication of Study 1with a few minor adjustments plus additional measures that were not reportedin this paper Again contrary to prediction remote influence scores were non-significantly higher in the block condition than in the co-operate conditionThis replicates the trend found in Study 1 failing to support hypothesis oneThese results suggest that the mental strategy adopted by the influencee in theface of a remote influence attempt has no effect on remote influence perfor-mance Referring back to the questions posed in the introduction then wehave not found any evidence to suggest that it is possible to resist a remote in-fluence effect and there is as yet no indication of the possible limiting condi-tions for remote mental influence

Unlike Study 1 Study 2 found statistically significant evidence of an overallremote influence effect in the prespecified direction (ie calming of EDAwhen that was the influencerrsquos intention activation of EDA when that was theinfluencerrsquos intention) In order to facilitate comparison with a recent meta-an-alytic review of remote influence studies we calculated the effect size (r) ac-cording to the formula r = [t2(t2+df)]12 For their review of 19 EDA influencestudies Schlitz and Braud (1997) found an average effect size of r = +25 OurStudy 1 with overall t = - 031 df = 31 gives a negligible effect size of r = - 006 Study 2 with overall t = 1806 df = 49 gives r = +25 Study 2 there-fore replicates the mean effect size found in EDA studies conducted by otherresearchers in other parapsychology laboratories Post hoc we can combinethe data of the two studies for greater statistical power Overall t = 1176 df = 81 giving a small effect size of r = +13

We can find no clear reason why Study 2 found an EDA influence effectwhile Study 1 did not A number of factors are worth considering 1) The

Blocking vs Co-operating 531

532 C Watt et al

significant result could be due to chance or undetected artifact 2) Study 2 im-plemented what we considered to be minor methodological improvements thatmight have removed some artifactual fluctuations from the EDA data that weobserved in study one 3) Each experimenter in Study 2 conducted three testingsessions with their participant By the third session in which the remote influ-ence measure was taken experimenters were on quite friendly terms with theirparticipants perhaps more so than in Study 1 where each participant came infor only one testing session 4) The participants in Study 2 were more represen-tative of the general population we felt than those in Study 1 Study 1 in-volved more undergraduate psychology students and associates of the experi-menters 5) The experimenters in study one were less experienced with theremote influence protocol than when they conducted Study 2

In addition to the formal hypotheses we reported a number of exploratorymeasures Study 1 found that influenceesrsquo self-reported vividness of visual im-agery appeared to be unrelated to the remote influence performance In bothstudies influenceesrsquo qualitative descriptions of the actual mental strategiesthey employed when blocking and co-operating revealed a tendency for differ-ent strategies to be preferred for the different conditions The use of visual im-agery was popular for both blocking (eg visualizing being surrounded by aprotective cocoon) and co-operating (eg visualizing a river flowing betweeninfluencer and influencee) However a considerable proportion of participantsreported adopting an open or relaxed frame of mind when co-operating whilstalmost no participants chose this strategy when blocking This was despite thefact that prior to each session we encouraged participants to adopt comparablestrategies for blocking and co-operating

It was interesting and perhaps surprising to note that in both studies themajority of participants reported no feelings of concern or discomfort at thepossibility of having their physiology remotely influenced Those who wereconcerned tended to be those who knew more about parapsychology and thepossible harmful applications of remote influence A small proportion of par-ticipants was unconcerned because they did not believe such an influence waspossible in the first place

Both studies employed an exploratory variance measure of EDA influenceThis measure showed a greater range of scoring overall thus perhaps greaterpotential sensitivity to a remote influence effect However the variance mea-sure showed no indication of a remote influence effect To date researchershave employed a variety of indicators of remote influence as the literaturegrows progress may be made on identifying which is the most appropriate in-dicator

Neither study found significant sex differences in remote influence effectsthough in both studies female influencees tended to have higher remote influ-ence scores than male influencees

In terms of their overall remote influence effects there appeared to be nodifference between experimenterinfluencers in Study 1 (r = +06 - 05 and

- 06) In Study 2 effect size measures indicated that the experimenters dif-fered in remote influence effect sizes (r = +17 and +30) Due to the smaller Nthis difference was not found to be statistically significant Post hoc combin-ing Studies 1 and 2 for experimenters JW and CW (ZMcD was not an experi-menter in Study 2) the effect sizes are r = +08 and r = +20 respectively

We conducted a post hoc analysis of the effects of influenceeinfluencer sexpairings on remote influence scores Our two studies came up with contradic-tory findings each statistically significant Study one found that mixed sexpairings had significantly higher remote influence scores than same sex pair-ings This replicated the post hoc finding of Schlitz and LaBerge (1997) in con-nection with the autonomic detection of remote observation Study 2 foundthat same sex pairings had significantly higher remote influence scores thanmixed sex pairings Our contradictory findings do not help to clarify the pic-ture on the importance or otherwise of sex pairings It is quite possible that sexis not the most salient aspect of the dynamic between pairs Age personalityand familiarity are likely to be important variables moderating the relationshipbetween influencer and influencee

There are many more variables that we could look at but as the number ofanalyses multiplies so does the opportunity for spurious significant resultsFor this reason we have named only two formal hypotheses for each experi-ment All other analyses are regarded as exploratory and any trends that wehave highlighted above would require replication

In conclusion these two studies have found no evidence of a difference inremote influence effects when influencees are asked to adopt cooperative men-tal strategies compared to when they are asked to adopt blocking mental strate-gies In fact contrary to expectation both studies found slightly higher remoteinfluence effects for the blocking condition compared to the co-operating con-dition On this basis we have found no indication of limiting conditions for re-mote influence effects Therefore while anecdotal advice exists on how to de-fend oneself against a remote physiological influence attempt in the realworld laboratory evidence to date has provided no support for this advice

Given the importance of this question both in practical terms and in theoret-ical terms we argue that this line of research deserves continued investigationIf a way could be found to conduct research with a more realistic scenario mdash where the remote influence was not just neutrally and abstractlydefined as unwelcome mdash then our research might have more meaningful thingsto say about the parameters of remote influence It is interesting and perhapssurprising to note that when Watt has lectured on this topic to informed lay au-diences the feedback has often been that researchers must use more potential-ly harmful or disruptive remote influence (of course with the informed consentof the participants) While there are clear ethical inhibitions when consideringpossible physiological influence perhaps research using higher level influencewould be more acceptable For example one could attempt to disrupt or en-hance the concentration of chess players who were aware of the influence

Blocking vs Co-operating 533

534 C Watt et al

attempt and who were concurrently blocking or co-operating with the influ-ence attempt Research already exists to suggest that remote influence does notjust operate at physiological levels Attention focusing for example can be fa-cilitated by a remote helper (Brady amp Morris 1997 Braud et al 1995) Futureresearch in this area must therefore expand beyond physiological influences

Acknowledgements

The authors gratefully acknowledge the financial support they receivedfrom the Institut fuumlr Grenzgebiete der Psychologie und PsychohygieneFreiburg Germany while conducting these studies We would also like tothank Dr Paul Stevens for generously helping to solve software difficultieswhen preparing for these studies and the anonymous reviewer for helpfulcomments Address reprint requests to Dr C Watt Department of PsychologyUniversity of Edinburgh 7 George Square Edinburgh EH8 9JZ Scotland UK

ReferencesBrady C amp Morris R L (1997) Attention focusing facilitated through remote mental interac-

tion A replication and exploration of parameters In R Wiseman (Ed) Proceedings of Pre-sented Papers The Parapsychological Association 40th Annual Convention held in conjunc-tion with The Society for Psychical Research pp 73ndash91

Braud W G (1985) Blockingshielding psychic functioning through psychological and psychictechniques A report of three preliminary studies In RA White and J Solfvin (Eds) Re-search in Parapsychology 1984 Metuchen NJ Scarecrow Press pp 42ndash44

Braud W G amp Schlitz M J (1991) Consciousness interactions with remote biological systemsAnomalous intentionality effects Subtle Energies 2 1

Braud W G Schlitz M J Collins J and Klitch H (1985) Further studies of the bio-PK effectFeedback blocking specificitygenerality In RA White amp J Solfvin (Eds) Research inParapsychology 1984 Metuchen NJ Scarecrow Press pp 45ndash48

Braud W G Shafer D McNeill K and Guerra V (1995) Attention focusing facilitatedthrough remote mental interaction Journal of the American Society for Psychical Research89 103

Cohen J (1997) Statistical Power Analysis for the Behavioral Sciences New York AcademicPress

Dalton K (1994) Remotely influenced ESP performance in a computer task A preliminarystudy Proceedings of the 37th Annual Convention of the Parapsychological Association pp95ndash103

Dossey L (1994) Healing and the mind Is there a dark side Journal of Scientific Exploration 873

Halifax-Grof J (1974) Hex death In A Angoff amp D Barth (Eds) Parapsychology and Anthro-pology Proceedings of an International Conference Held in London England August 29ndash311973 NY Parapsychology Foundation

Hope M (1983) Practical Techniques of Psychic Self-Defence Wellingborough Northampton-shire Aquarian Press

Marks DF (1973) Visual imagery differences in the recall of pictures British Journal of Psy-chology 64 17

Nash CB (1982) Psychokinetic control of bacterial growth Journal of the Society for PsychicalResearch 51 217

Radin DI Taylor RK and Braud WG (1995) Remote mental influence of human electroder-mal activity A pilot replication European Journal of Parapsychology 11 19

Schlitz M amp Braud W G (1997) Distant intentionality and healing Assessing the evidence Al-ternative Therapies 3 62

Schlitz M J amp LaBerge S (1997) Covert observation increases skin conductance in subjectsunaware of when they are being observed A replication Journal of Parapsychology 61 185

Schmeidler G R (1958) Agent-percipient relationships Journal of the American Society forPsychical Research 52 47

Schmeidler G R (1961) Evidence for two kinds of telepathy International Journal of Parapsy-chology 3 5

Tart C T (1979) A survey of expert opinion on potentially negative uses of psi United Statesgovernment interest in psi and the level of research funding of the field In WG Roll (Ed) Re-search in Parapsychology 1978 Metuchen NJ Scarecrow Press pp 54

Tart C T amp Labore C M (1986) Attitudes towards strongly functioning psi A preliminary sur-vey Journal of the America Society for Psychical Research 80 163

Tart C T (1986) Psychicsrsquo fears of psychic powers Journal of the American Society for Psychi-cal Research 80 278

Watt C A McDermott Z amp Ravenscroft J (1997) Exploring the limits of remote mental inter-action A comparison of blocking versus co-operating strategies In R Wiseman (Ed) Pro-ceedings of the 40th Annual Convention of the Parapsychological Association held in conjunc-tion with the Society for Psychical Research Brighton England pp 464ndash476

Blocking vs Co-operating 535

to settle down However in order to remove any remaining potential artifactu-al effects on EDA in Study 2 only visual instructions were given

Another difference in methodology concerned the screensaver viewed bythe participants during the experimental session In Study 1 some participantsreported that the series of patterns produced by the computer were being dis-played at too fast a rate such that some participants found it distracting or dis-turbing Again we felt that this may have an artifactual effect on EDA so wedecided for study two to reduce the speed of the screensaver

Thirdly in Study 2 we did not give participants detailed information abouttheir remote influence performance until the feedback letters that were distrib-uted at the conclusion of the entire study The reason for this change was mere-ly to streamline the procedure on what was already a time-consuming experi-ment Finally in another streamlining change we did not take VVIQ measuresin Study 2

Planned Hypotheses and Analyses

These were the same as with Study 1 That is hypothesis 1 predicted higherPSI scores for the co-operating condition than for the blocking condition(measured with a related t-test one-tailed) If no significant difference wasfound between blocking PSI and co-operating PSI scores the two would becombined to look at overall remote influence effect (hypothesis 2 single-meant-test one-tailed) If a significant difference was found with hypothesis 1 theanalysis for hypothesis 2 would be conducted with the data for co-operatingsessions only

Exploratory Questions

As with study one we took an exploratory variance measure of remote EDAinfluence We also gathered qualitative descriptions of the mental strategiesparticipants adopted whilst blocking and co-operating We also looked at sexeffects experimenter effects and post hoc influenceeinfluencer sex pairings

Participants

Study 2 had two experimenterinfluencers (Watt amp Ravenscroft) After aparticipant was recruited (usually by Ravenscroft) they were assigned to oneof the two experimenters for all three testing sessions This assignment wasdone unsystematically according to the preferences and availability of the ex-perimenters As planned 50 volunteer participants were drawn from the gener-al participant pool at the Koestler Parapsychology Unit and by word ofmouth All participants completed the Participant Information Form and aConsent form prior to the study The participants consisted of 33 females and17 males Average age was 41 years range from 22 years to 76 years of age Inorder to increase the studyrsquos relevance to the general population undergradu-ate psychology students were not recruited as participants The main criterion

Blocking vs Co-operating 527

528 C Watt et al

for participation was that the participants were interested in and open-mindedabout parapsychology Again there were no drop-outs from this study Allparticipants received a detailed personal feedback letter at the conclusion ofthe study

Study Two Results and Discussion

Remote Influence Measure

PSI scores are summarized in Table 6 which shows that the mean PSI scorewhen co-operating was nonsignificantly lower than mean PSI score whenblocking (t = - 595 df = 49) Hypothesis 1 was therefore not supported Asplanned in the absence of any significant difference between blocking and co-operating PSI scores the two conditions were combined to test hypothesis 2which predicted overall significant remote mental influence performance Thecombined PSI score was significantly greater than mean chance expectation of50 (t = 1806 df = 49 p = 038 1 - t) Therefore hypothesis 2 was supportedgiving evidence of an overall remote influence effect upon psychophysiologyin the prespecified direction

The failure to support hypothesis 1 confirms what we found in Study 1 thatthere was no significant difference between blocking and co-operating condi-tions on remote influence scores Indeed both studies found a slight effect inthe opposite direction to that predicted that is slightly higher PSI scores whenblocking was the influenceersquos intention than when co-operating was their in-tention It was encouraging to find significant support for a remote influenceeffect in the predicted direction for blocking and co-operating conditions com-bined No such trend was found in the previous study though with a smallernumber of participants (N = 32) the earlier study had lower statistical powerthan the present study

Exploratory Questions

Variance Measure

Table 7 shows the results for the exploratory measure of EDA variance lt 50represents reduced variance gt 50 represents increased variance We found nodifference between variance scores for cooperate versus block conditions(t = 150 df = 49) When these are combined the resulting figure does not de-viate from chance expectation (t = 440 df = 49) Thus as in Study 1 it seems

TABLE 6Summary Data Mean PSI Scores

(N = 50 MCE = 50 range in brackets)

Co-operate PSI Block PSI Combined PSI

5008 5019 5016(4821- 5221) (4807- 5557) (4900- 5296)

that the exploratory variance measure does not reveal any additional indica-tion of the operation of a remote mental influence in this study However thewider range of variance scoring continues to suggest that this measure mayhave greater potential sensitivity than the summary PSI score

Influenceesrsquo Blocking and Co-operating Strategies

We were interested in learning more about what particular strategies partici-pants chose to adopt when blocking and co-operating As in Study 1 a post-session questionnaire was used to gather this information Table 8 shows thepercentage of participants who adopted each kind of strategy

The table shows a very similar pattern to that found in Study 1 The mainpoint of interest is that almost no participants took a passive approach toblocking (ie adopting an openrelaxing strategy) A much higher proportionpreferred to adopt this approach when co-operating The preferred mode forboth conditions was to use some form of visual imagery such as imagining ashield or a screen when blocking and imagining an open channel such as ariver when co-operating

The majority of participants reported no concerns over the possibility ofhaving their physiology remotely influenced by another person as found instudy one

Sex Effects

As in study one the sexes were not equally represented among participantsWhereas Study 1 had an excess of male participants in Study 2 twice as manyfemales (N = 33) as males (N = 17) participated It was therefore important tolook for sex differences since any such differences could confound the inter-pretation of overall trends We found no difference between sexes when co-op-erating (female mean PSI score = 5008 male mean PSI score = 5006 t = 107 df = 48) There was a nonsignificant trend for females to have higher

Blocking vs Co-operating 529

TABLE 7Summary Data Mean Variance Scores(N = 50 MCE = 50 range in brackets)

Co-operate Variance Block Variance Combined Variance

5076 5054 5034(3164- 7279) (3572- 6813) (3749- 6671)

TABLE 8Influenceesrsquo Mental Strategies ()

Visual Imagery Cognitive VisCog OpenRelax

Blocking 64 26 6 2Co-operating 60 8 12 20

530 C Watt et al

PSI scores than males when blocking (mean PSI scores 5030 and 4998 re-spectively t = 992 df = 48) Combining blocking and co-operating conditionsthere was no overall difference between the sexes (mean PSI scores 5020 forfemales 5008 for males t = 665 df = 48)

Experimenter Effects

Study 2 found a statistically significant overall remote influence effect(combined PSI = 50155 t = 1806) Table 9 shows the breakdown of remoteinfluence scores for the two experimenters CW and JR Looking at the overalldata for the two experimenters we see that CWrsquos sessions were associatedwith most of the remote influence effect However due to the smaller N CWrsquossessions are not independently significant (t = 1632) and there is not a signifi-cant difference between CW and JRrsquos overall PSI scores (t = 678) Post hocwe calculated effect size r according to the formula r = [t2(t2+df)]12 for the twoexperimenters For JR (t = 817 df = 22) we found a small effect sizer = +17 For CW (t = 1632 df = 26) the effect size was moderate r = +30This may be some post hoc indication of an experimenter effect in Study 2

InfluenceeInfluencer Sex Pairings

As can be seen from Table 10 Study 2 found exactly the opposite trend tothat found in Study 1 On a post hoc test same sex pairings scored significant-ly higher on the remote influence PSI measure than mixed sex pairings(t = - 2034 df = 48 p = 048 2 - t) This contradicts Schlitz and LaBergersquos(1997) finding of greater autonomic detection of remote staring for mixed sexpairs

TABLE 9Remote Influence Results by Experimenter

Experimenter N sessions Coop PSI Block PSI Combined PSI

CW 27 5011 5026 5021JR 23 5003 5010 5009

TABLE 10Mean PSI Scores for Different InfluenceeInfluencer Sex Pairings

Pairing N Coop PSI Block PSI Combined PSI

MF 11 4977 5009 4997FM 17 4983 5021 5003FF 16 5035 5038 5038MM 6 5060 4978 5027

General Summary and Conclusions

We have described the findings of two studies exploring the limits of remotemental influence effects We wished to see whether the mental strategiesadopted by the influencee would affect remote influence scores In particularcould an influencee block or resist a remote influence attempt We recorded in-fluenceesrsquo electrodermal activity while the influencer attempted to calm to ac-tivate or to have no influence over the influenceesrsquo physiology In each studythere were two conditions reflecting the influenceesrsquo designated mental strate-gies in response to the remote influence attempt blocking and co-operating

There were only two formal hypotheses 1) that remote influence scoreswould be significantly higher during the co-operating condition than duringthe blocking condition and 2) that overall there would be significant evidenceof remote mental influence on physiology in the designated direction (iecalming when calming was the influencerrsquos intention activating when activat-ing was the influencerrsquos intention) Contrary to prediction Study 1 found thatremote influence effects were higher in the block condition than in the co-op-erate condition (but not to a significant extent) Also overall remote influenceperformance was at chance level This study therefore found no support for ei-ther of the formal hypotheses Study 2 consisted of a replication of Study 1with a few minor adjustments plus additional measures that were not reportedin this paper Again contrary to prediction remote influence scores were non-significantly higher in the block condition than in the co-operate conditionThis replicates the trend found in Study 1 failing to support hypothesis oneThese results suggest that the mental strategy adopted by the influencee in theface of a remote influence attempt has no effect on remote influence perfor-mance Referring back to the questions posed in the introduction then wehave not found any evidence to suggest that it is possible to resist a remote in-fluence effect and there is as yet no indication of the possible limiting condi-tions for remote mental influence

Unlike Study 1 Study 2 found statistically significant evidence of an overallremote influence effect in the prespecified direction (ie calming of EDAwhen that was the influencerrsquos intention activation of EDA when that was theinfluencerrsquos intention) In order to facilitate comparison with a recent meta-an-alytic review of remote influence studies we calculated the effect size (r) ac-cording to the formula r = [t2(t2+df)]12 For their review of 19 EDA influencestudies Schlitz and Braud (1997) found an average effect size of r = +25 OurStudy 1 with overall t = - 031 df = 31 gives a negligible effect size of r = - 006 Study 2 with overall t = 1806 df = 49 gives r = +25 Study 2 there-fore replicates the mean effect size found in EDA studies conducted by otherresearchers in other parapsychology laboratories Post hoc we can combinethe data of the two studies for greater statistical power Overall t = 1176 df = 81 giving a small effect size of r = +13

We can find no clear reason why Study 2 found an EDA influence effectwhile Study 1 did not A number of factors are worth considering 1) The

Blocking vs Co-operating 531

532 C Watt et al

significant result could be due to chance or undetected artifact 2) Study 2 im-plemented what we considered to be minor methodological improvements thatmight have removed some artifactual fluctuations from the EDA data that weobserved in study one 3) Each experimenter in Study 2 conducted three testingsessions with their participant By the third session in which the remote influ-ence measure was taken experimenters were on quite friendly terms with theirparticipants perhaps more so than in Study 1 where each participant came infor only one testing session 4) The participants in Study 2 were more represen-tative of the general population we felt than those in Study 1 Study 1 in-volved more undergraduate psychology students and associates of the experi-menters 5) The experimenters in study one were less experienced with theremote influence protocol than when they conducted Study 2

In addition to the formal hypotheses we reported a number of exploratorymeasures Study 1 found that influenceesrsquo self-reported vividness of visual im-agery appeared to be unrelated to the remote influence performance In bothstudies influenceesrsquo qualitative descriptions of the actual mental strategiesthey employed when blocking and co-operating revealed a tendency for differ-ent strategies to be preferred for the different conditions The use of visual im-agery was popular for both blocking (eg visualizing being surrounded by aprotective cocoon) and co-operating (eg visualizing a river flowing betweeninfluencer and influencee) However a considerable proportion of participantsreported adopting an open or relaxed frame of mind when co-operating whilstalmost no participants chose this strategy when blocking This was despite thefact that prior to each session we encouraged participants to adopt comparablestrategies for blocking and co-operating

It was interesting and perhaps surprising to note that in both studies themajority of participants reported no feelings of concern or discomfort at thepossibility of having their physiology remotely influenced Those who wereconcerned tended to be those who knew more about parapsychology and thepossible harmful applications of remote influence A small proportion of par-ticipants was unconcerned because they did not believe such an influence waspossible in the first place

Both studies employed an exploratory variance measure of EDA influenceThis measure showed a greater range of scoring overall thus perhaps greaterpotential sensitivity to a remote influence effect However the variance mea-sure showed no indication of a remote influence effect To date researchershave employed a variety of indicators of remote influence as the literaturegrows progress may be made on identifying which is the most appropriate in-dicator

Neither study found significant sex differences in remote influence effectsthough in both studies female influencees tended to have higher remote influ-ence scores than male influencees

In terms of their overall remote influence effects there appeared to be nodifference between experimenterinfluencers in Study 1 (r = +06 - 05 and

- 06) In Study 2 effect size measures indicated that the experimenters dif-fered in remote influence effect sizes (r = +17 and +30) Due to the smaller Nthis difference was not found to be statistically significant Post hoc combin-ing Studies 1 and 2 for experimenters JW and CW (ZMcD was not an experi-menter in Study 2) the effect sizes are r = +08 and r = +20 respectively

We conducted a post hoc analysis of the effects of influenceeinfluencer sexpairings on remote influence scores Our two studies came up with contradic-tory findings each statistically significant Study one found that mixed sexpairings had significantly higher remote influence scores than same sex pair-ings This replicated the post hoc finding of Schlitz and LaBerge (1997) in con-nection with the autonomic detection of remote observation Study 2 foundthat same sex pairings had significantly higher remote influence scores thanmixed sex pairings Our contradictory findings do not help to clarify the pic-ture on the importance or otherwise of sex pairings It is quite possible that sexis not the most salient aspect of the dynamic between pairs Age personalityand familiarity are likely to be important variables moderating the relationshipbetween influencer and influencee

There are many more variables that we could look at but as the number ofanalyses multiplies so does the opportunity for spurious significant resultsFor this reason we have named only two formal hypotheses for each experi-ment All other analyses are regarded as exploratory and any trends that wehave highlighted above would require replication

In conclusion these two studies have found no evidence of a difference inremote influence effects when influencees are asked to adopt cooperative men-tal strategies compared to when they are asked to adopt blocking mental strate-gies In fact contrary to expectation both studies found slightly higher remoteinfluence effects for the blocking condition compared to the co-operating con-dition On this basis we have found no indication of limiting conditions for re-mote influence effects Therefore while anecdotal advice exists on how to de-fend oneself against a remote physiological influence attempt in the realworld laboratory evidence to date has provided no support for this advice

Given the importance of this question both in practical terms and in theoret-ical terms we argue that this line of research deserves continued investigationIf a way could be found to conduct research with a more realistic scenario mdash where the remote influence was not just neutrally and abstractlydefined as unwelcome mdash then our research might have more meaningful thingsto say about the parameters of remote influence It is interesting and perhapssurprising to note that when Watt has lectured on this topic to informed lay au-diences the feedback has often been that researchers must use more potential-ly harmful or disruptive remote influence (of course with the informed consentof the participants) While there are clear ethical inhibitions when consideringpossible physiological influence perhaps research using higher level influencewould be more acceptable For example one could attempt to disrupt or en-hance the concentration of chess players who were aware of the influence

Blocking vs Co-operating 533

534 C Watt et al

attempt and who were concurrently blocking or co-operating with the influ-ence attempt Research already exists to suggest that remote influence does notjust operate at physiological levels Attention focusing for example can be fa-cilitated by a remote helper (Brady amp Morris 1997 Braud et al 1995) Futureresearch in this area must therefore expand beyond physiological influences

Acknowledgements

The authors gratefully acknowledge the financial support they receivedfrom the Institut fuumlr Grenzgebiete der Psychologie und PsychohygieneFreiburg Germany while conducting these studies We would also like tothank Dr Paul Stevens for generously helping to solve software difficultieswhen preparing for these studies and the anonymous reviewer for helpfulcomments Address reprint requests to Dr C Watt Department of PsychologyUniversity of Edinburgh 7 George Square Edinburgh EH8 9JZ Scotland UK

ReferencesBrady C amp Morris R L (1997) Attention focusing facilitated through remote mental interac-

tion A replication and exploration of parameters In R Wiseman (Ed) Proceedings of Pre-sented Papers The Parapsychological Association 40th Annual Convention held in conjunc-tion with The Society for Psychical Research pp 73ndash91

Braud W G (1985) Blockingshielding psychic functioning through psychological and psychictechniques A report of three preliminary studies In RA White and J Solfvin (Eds) Re-search in Parapsychology 1984 Metuchen NJ Scarecrow Press pp 42ndash44

Braud W G amp Schlitz M J (1991) Consciousness interactions with remote biological systemsAnomalous intentionality effects Subtle Energies 2 1

Braud W G Schlitz M J Collins J and Klitch H (1985) Further studies of the bio-PK effectFeedback blocking specificitygenerality In RA White amp J Solfvin (Eds) Research inParapsychology 1984 Metuchen NJ Scarecrow Press pp 45ndash48

Braud W G Shafer D McNeill K and Guerra V (1995) Attention focusing facilitatedthrough remote mental interaction Journal of the American Society for Psychical Research89 103

Cohen J (1997) Statistical Power Analysis for the Behavioral Sciences New York AcademicPress

Dalton K (1994) Remotely influenced ESP performance in a computer task A preliminarystudy Proceedings of the 37th Annual Convention of the Parapsychological Association pp95ndash103

Dossey L (1994) Healing and the mind Is there a dark side Journal of Scientific Exploration 873

Halifax-Grof J (1974) Hex death In A Angoff amp D Barth (Eds) Parapsychology and Anthro-pology Proceedings of an International Conference Held in London England August 29ndash311973 NY Parapsychology Foundation

Hope M (1983) Practical Techniques of Psychic Self-Defence Wellingborough Northampton-shire Aquarian Press

Marks DF (1973) Visual imagery differences in the recall of pictures British Journal of Psy-chology 64 17

Nash CB (1982) Psychokinetic control of bacterial growth Journal of the Society for PsychicalResearch 51 217

Radin DI Taylor RK and Braud WG (1995) Remote mental influence of human electroder-mal activity A pilot replication European Journal of Parapsychology 11 19

Schlitz M amp Braud W G (1997) Distant intentionality and healing Assessing the evidence Al-ternative Therapies 3 62

Schlitz M J amp LaBerge S (1997) Covert observation increases skin conductance in subjectsunaware of when they are being observed A replication Journal of Parapsychology 61 185

Schmeidler G R (1958) Agent-percipient relationships Journal of the American Society forPsychical Research 52 47

Schmeidler G R (1961) Evidence for two kinds of telepathy International Journal of Parapsy-chology 3 5

Tart C T (1979) A survey of expert opinion on potentially negative uses of psi United Statesgovernment interest in psi and the level of research funding of the field In WG Roll (Ed) Re-search in Parapsychology 1978 Metuchen NJ Scarecrow Press pp 54

Tart C T amp Labore C M (1986) Attitudes towards strongly functioning psi A preliminary sur-vey Journal of the America Society for Psychical Research 80 163

Tart C T (1986) Psychicsrsquo fears of psychic powers Journal of the American Society for Psychi-cal Research 80 278

Watt C A McDermott Z amp Ravenscroft J (1997) Exploring the limits of remote mental inter-action A comparison of blocking versus co-operating strategies In R Wiseman (Ed) Pro-ceedings of the 40th Annual Convention of the Parapsychological Association held in conjunc-tion with the Society for Psychical Research Brighton England pp 464ndash476

Blocking vs Co-operating 535

528 C Watt et al

for participation was that the participants were interested in and open-mindedabout parapsychology Again there were no drop-outs from this study Allparticipants received a detailed personal feedback letter at the conclusion ofthe study

Study Two Results and Discussion

Remote Influence Measure

PSI scores are summarized in Table 6 which shows that the mean PSI scorewhen co-operating was nonsignificantly lower than mean PSI score whenblocking (t = - 595 df = 49) Hypothesis 1 was therefore not supported Asplanned in the absence of any significant difference between blocking and co-operating PSI scores the two conditions were combined to test hypothesis 2which predicted overall significant remote mental influence performance Thecombined PSI score was significantly greater than mean chance expectation of50 (t = 1806 df = 49 p = 038 1 - t) Therefore hypothesis 2 was supportedgiving evidence of an overall remote influence effect upon psychophysiologyin the prespecified direction

The failure to support hypothesis 1 confirms what we found in Study 1 thatthere was no significant difference between blocking and co-operating condi-tions on remote influence scores Indeed both studies found a slight effect inthe opposite direction to that predicted that is slightly higher PSI scores whenblocking was the influenceersquos intention than when co-operating was their in-tention It was encouraging to find significant support for a remote influenceeffect in the predicted direction for blocking and co-operating conditions com-bined No such trend was found in the previous study though with a smallernumber of participants (N = 32) the earlier study had lower statistical powerthan the present study

Exploratory Questions

Variance Measure

Table 7 shows the results for the exploratory measure of EDA variance lt 50represents reduced variance gt 50 represents increased variance We found nodifference between variance scores for cooperate versus block conditions(t = 150 df = 49) When these are combined the resulting figure does not de-viate from chance expectation (t = 440 df = 49) Thus as in Study 1 it seems

TABLE 6Summary Data Mean PSI Scores

(N = 50 MCE = 50 range in brackets)

Co-operate PSI Block PSI Combined PSI

5008 5019 5016(4821- 5221) (4807- 5557) (4900- 5296)

that the exploratory variance measure does not reveal any additional indica-tion of the operation of a remote mental influence in this study However thewider range of variance scoring continues to suggest that this measure mayhave greater potential sensitivity than the summary PSI score

Influenceesrsquo Blocking and Co-operating Strategies

We were interested in learning more about what particular strategies partici-pants chose to adopt when blocking and co-operating As in Study 1 a post-session questionnaire was used to gather this information Table 8 shows thepercentage of participants who adopted each kind of strategy

The table shows a very similar pattern to that found in Study 1 The mainpoint of interest is that almost no participants took a passive approach toblocking (ie adopting an openrelaxing strategy) A much higher proportionpreferred to adopt this approach when co-operating The preferred mode forboth conditions was to use some form of visual imagery such as imagining ashield or a screen when blocking and imagining an open channel such as ariver when co-operating

The majority of participants reported no concerns over the possibility ofhaving their physiology remotely influenced by another person as found instudy one

Sex Effects

As in study one the sexes were not equally represented among participantsWhereas Study 1 had an excess of male participants in Study 2 twice as manyfemales (N = 33) as males (N = 17) participated It was therefore important tolook for sex differences since any such differences could confound the inter-pretation of overall trends We found no difference between sexes when co-op-erating (female mean PSI score = 5008 male mean PSI score = 5006 t = 107 df = 48) There was a nonsignificant trend for females to have higher

Blocking vs Co-operating 529

TABLE 7Summary Data Mean Variance Scores(N = 50 MCE = 50 range in brackets)

Co-operate Variance Block Variance Combined Variance

5076 5054 5034(3164- 7279) (3572- 6813) (3749- 6671)

TABLE 8Influenceesrsquo Mental Strategies ()

Visual Imagery Cognitive VisCog OpenRelax

Blocking 64 26 6 2Co-operating 60 8 12 20

530 C Watt et al

PSI scores than males when blocking (mean PSI scores 5030 and 4998 re-spectively t = 992 df = 48) Combining blocking and co-operating conditionsthere was no overall difference between the sexes (mean PSI scores 5020 forfemales 5008 for males t = 665 df = 48)

Experimenter Effects

Study 2 found a statistically significant overall remote influence effect(combined PSI = 50155 t = 1806) Table 9 shows the breakdown of remoteinfluence scores for the two experimenters CW and JR Looking at the overalldata for the two experimenters we see that CWrsquos sessions were associatedwith most of the remote influence effect However due to the smaller N CWrsquossessions are not independently significant (t = 1632) and there is not a signifi-cant difference between CW and JRrsquos overall PSI scores (t = 678) Post hocwe calculated effect size r according to the formula r = [t2(t2+df)]12 for the twoexperimenters For JR (t = 817 df = 22) we found a small effect sizer = +17 For CW (t = 1632 df = 26) the effect size was moderate r = +30This may be some post hoc indication of an experimenter effect in Study 2

InfluenceeInfluencer Sex Pairings

As can be seen from Table 10 Study 2 found exactly the opposite trend tothat found in Study 1 On a post hoc test same sex pairings scored significant-ly higher on the remote influence PSI measure than mixed sex pairings(t = - 2034 df = 48 p = 048 2 - t) This contradicts Schlitz and LaBergersquos(1997) finding of greater autonomic detection of remote staring for mixed sexpairs

TABLE 9Remote Influence Results by Experimenter

Experimenter N sessions Coop PSI Block PSI Combined PSI

CW 27 5011 5026 5021JR 23 5003 5010 5009

TABLE 10Mean PSI Scores for Different InfluenceeInfluencer Sex Pairings

Pairing N Coop PSI Block PSI Combined PSI

MF 11 4977 5009 4997FM 17 4983 5021 5003FF 16 5035 5038 5038MM 6 5060 4978 5027

General Summary and Conclusions

We have described the findings of two studies exploring the limits of remotemental influence effects We wished to see whether the mental strategiesadopted by the influencee would affect remote influence scores In particularcould an influencee block or resist a remote influence attempt We recorded in-fluenceesrsquo electrodermal activity while the influencer attempted to calm to ac-tivate or to have no influence over the influenceesrsquo physiology In each studythere were two conditions reflecting the influenceesrsquo designated mental strate-gies in response to the remote influence attempt blocking and co-operating

There were only two formal hypotheses 1) that remote influence scoreswould be significantly higher during the co-operating condition than duringthe blocking condition and 2) that overall there would be significant evidenceof remote mental influence on physiology in the designated direction (iecalming when calming was the influencerrsquos intention activating when activat-ing was the influencerrsquos intention) Contrary to prediction Study 1 found thatremote influence effects were higher in the block condition than in the co-op-erate condition (but not to a significant extent) Also overall remote influenceperformance was at chance level This study therefore found no support for ei-ther of the formal hypotheses Study 2 consisted of a replication of Study 1with a few minor adjustments plus additional measures that were not reportedin this paper Again contrary to prediction remote influence scores were non-significantly higher in the block condition than in the co-operate conditionThis replicates the trend found in Study 1 failing to support hypothesis oneThese results suggest that the mental strategy adopted by the influencee in theface of a remote influence attempt has no effect on remote influence perfor-mance Referring back to the questions posed in the introduction then wehave not found any evidence to suggest that it is possible to resist a remote in-fluence effect and there is as yet no indication of the possible limiting condi-tions for remote mental influence

Unlike Study 1 Study 2 found statistically significant evidence of an overallremote influence effect in the prespecified direction (ie calming of EDAwhen that was the influencerrsquos intention activation of EDA when that was theinfluencerrsquos intention) In order to facilitate comparison with a recent meta-an-alytic review of remote influence studies we calculated the effect size (r) ac-cording to the formula r = [t2(t2+df)]12 For their review of 19 EDA influencestudies Schlitz and Braud (1997) found an average effect size of r = +25 OurStudy 1 with overall t = - 031 df = 31 gives a negligible effect size of r = - 006 Study 2 with overall t = 1806 df = 49 gives r = +25 Study 2 there-fore replicates the mean effect size found in EDA studies conducted by otherresearchers in other parapsychology laboratories Post hoc we can combinethe data of the two studies for greater statistical power Overall t = 1176 df = 81 giving a small effect size of r = +13

We can find no clear reason why Study 2 found an EDA influence effectwhile Study 1 did not A number of factors are worth considering 1) The

Blocking vs Co-operating 531

532 C Watt et al

significant result could be due to chance or undetected artifact 2) Study 2 im-plemented what we considered to be minor methodological improvements thatmight have removed some artifactual fluctuations from the EDA data that weobserved in study one 3) Each experimenter in Study 2 conducted three testingsessions with their participant By the third session in which the remote influ-ence measure was taken experimenters were on quite friendly terms with theirparticipants perhaps more so than in Study 1 where each participant came infor only one testing session 4) The participants in Study 2 were more represen-tative of the general population we felt than those in Study 1 Study 1 in-volved more undergraduate psychology students and associates of the experi-menters 5) The experimenters in study one were less experienced with theremote influence protocol than when they conducted Study 2

In addition to the formal hypotheses we reported a number of exploratorymeasures Study 1 found that influenceesrsquo self-reported vividness of visual im-agery appeared to be unrelated to the remote influence performance In bothstudies influenceesrsquo qualitative descriptions of the actual mental strategiesthey employed when blocking and co-operating revealed a tendency for differ-ent strategies to be preferred for the different conditions The use of visual im-agery was popular for both blocking (eg visualizing being surrounded by aprotective cocoon) and co-operating (eg visualizing a river flowing betweeninfluencer and influencee) However a considerable proportion of participantsreported adopting an open or relaxed frame of mind when co-operating whilstalmost no participants chose this strategy when blocking This was despite thefact that prior to each session we encouraged participants to adopt comparablestrategies for blocking and co-operating

It was interesting and perhaps surprising to note that in both studies themajority of participants reported no feelings of concern or discomfort at thepossibility of having their physiology remotely influenced Those who wereconcerned tended to be those who knew more about parapsychology and thepossible harmful applications of remote influence A small proportion of par-ticipants was unconcerned because they did not believe such an influence waspossible in the first place

Both studies employed an exploratory variance measure of EDA influenceThis measure showed a greater range of scoring overall thus perhaps greaterpotential sensitivity to a remote influence effect However the variance mea-sure showed no indication of a remote influence effect To date researchershave employed a variety of indicators of remote influence as the literaturegrows progress may be made on identifying which is the most appropriate in-dicator

Neither study found significant sex differences in remote influence effectsthough in both studies female influencees tended to have higher remote influ-ence scores than male influencees

In terms of their overall remote influence effects there appeared to be nodifference between experimenterinfluencers in Study 1 (r = +06 - 05 and

- 06) In Study 2 effect size measures indicated that the experimenters dif-fered in remote influence effect sizes (r = +17 and +30) Due to the smaller Nthis difference was not found to be statistically significant Post hoc combin-ing Studies 1 and 2 for experimenters JW and CW (ZMcD was not an experi-menter in Study 2) the effect sizes are r = +08 and r = +20 respectively

We conducted a post hoc analysis of the effects of influenceeinfluencer sexpairings on remote influence scores Our two studies came up with contradic-tory findings each statistically significant Study one found that mixed sexpairings had significantly higher remote influence scores than same sex pair-ings This replicated the post hoc finding of Schlitz and LaBerge (1997) in con-nection with the autonomic detection of remote observation Study 2 foundthat same sex pairings had significantly higher remote influence scores thanmixed sex pairings Our contradictory findings do not help to clarify the pic-ture on the importance or otherwise of sex pairings It is quite possible that sexis not the most salient aspect of the dynamic between pairs Age personalityand familiarity are likely to be important variables moderating the relationshipbetween influencer and influencee

There are many more variables that we could look at but as the number ofanalyses multiplies so does the opportunity for spurious significant resultsFor this reason we have named only two formal hypotheses for each experi-ment All other analyses are regarded as exploratory and any trends that wehave highlighted above would require replication

In conclusion these two studies have found no evidence of a difference inremote influence effects when influencees are asked to adopt cooperative men-tal strategies compared to when they are asked to adopt blocking mental strate-gies In fact contrary to expectation both studies found slightly higher remoteinfluence effects for the blocking condition compared to the co-operating con-dition On this basis we have found no indication of limiting conditions for re-mote influence effects Therefore while anecdotal advice exists on how to de-fend oneself against a remote physiological influence attempt in the realworld laboratory evidence to date has provided no support for this advice

Given the importance of this question both in practical terms and in theoret-ical terms we argue that this line of research deserves continued investigationIf a way could be found to conduct research with a more realistic scenario mdash where the remote influence was not just neutrally and abstractlydefined as unwelcome mdash then our research might have more meaningful thingsto say about the parameters of remote influence It is interesting and perhapssurprising to note that when Watt has lectured on this topic to informed lay au-diences the feedback has often been that researchers must use more potential-ly harmful or disruptive remote influence (of course with the informed consentof the participants) While there are clear ethical inhibitions when consideringpossible physiological influence perhaps research using higher level influencewould be more acceptable For example one could attempt to disrupt or en-hance the concentration of chess players who were aware of the influence

Blocking vs Co-operating 533

534 C Watt et al

attempt and who were concurrently blocking or co-operating with the influ-ence attempt Research already exists to suggest that remote influence does notjust operate at physiological levels Attention focusing for example can be fa-cilitated by a remote helper (Brady amp Morris 1997 Braud et al 1995) Futureresearch in this area must therefore expand beyond physiological influences

Acknowledgements

The authors gratefully acknowledge the financial support they receivedfrom the Institut fuumlr Grenzgebiete der Psychologie und PsychohygieneFreiburg Germany while conducting these studies We would also like tothank Dr Paul Stevens for generously helping to solve software difficultieswhen preparing for these studies and the anonymous reviewer for helpfulcomments Address reprint requests to Dr C Watt Department of PsychologyUniversity of Edinburgh 7 George Square Edinburgh EH8 9JZ Scotland UK

ReferencesBrady C amp Morris R L (1997) Attention focusing facilitated through remote mental interac-

tion A replication and exploration of parameters In R Wiseman (Ed) Proceedings of Pre-sented Papers The Parapsychological Association 40th Annual Convention held in conjunc-tion with The Society for Psychical Research pp 73ndash91

Braud W G (1985) Blockingshielding psychic functioning through psychological and psychictechniques A report of three preliminary studies In RA White and J Solfvin (Eds) Re-search in Parapsychology 1984 Metuchen NJ Scarecrow Press pp 42ndash44

Braud W G amp Schlitz M J (1991) Consciousness interactions with remote biological systemsAnomalous intentionality effects Subtle Energies 2 1

Braud W G Schlitz M J Collins J and Klitch H (1985) Further studies of the bio-PK effectFeedback blocking specificitygenerality In RA White amp J Solfvin (Eds) Research inParapsychology 1984 Metuchen NJ Scarecrow Press pp 45ndash48

Braud W G Shafer D McNeill K and Guerra V (1995) Attention focusing facilitatedthrough remote mental interaction Journal of the American Society for Psychical Research89 103

Cohen J (1997) Statistical Power Analysis for the Behavioral Sciences New York AcademicPress

Dalton K (1994) Remotely influenced ESP performance in a computer task A preliminarystudy Proceedings of the 37th Annual Convention of the Parapsychological Association pp95ndash103

Dossey L (1994) Healing and the mind Is there a dark side Journal of Scientific Exploration 873

Halifax-Grof J (1974) Hex death In A Angoff amp D Barth (Eds) Parapsychology and Anthro-pology Proceedings of an International Conference Held in London England August 29ndash311973 NY Parapsychology Foundation

Hope M (1983) Practical Techniques of Psychic Self-Defence Wellingborough Northampton-shire Aquarian Press

Marks DF (1973) Visual imagery differences in the recall of pictures British Journal of Psy-chology 64 17

Nash CB (1982) Psychokinetic control of bacterial growth Journal of the Society for PsychicalResearch 51 217

Radin DI Taylor RK and Braud WG (1995) Remote mental influence of human electroder-mal activity A pilot replication European Journal of Parapsychology 11 19

Schlitz M amp Braud W G (1997) Distant intentionality and healing Assessing the evidence Al-ternative Therapies 3 62

Schlitz M J amp LaBerge S (1997) Covert observation increases skin conductance in subjectsunaware of when they are being observed A replication Journal of Parapsychology 61 185

Schmeidler G R (1958) Agent-percipient relationships Journal of the American Society forPsychical Research 52 47

Schmeidler G R (1961) Evidence for two kinds of telepathy International Journal of Parapsy-chology 3 5

Tart C T (1979) A survey of expert opinion on potentially negative uses of psi United Statesgovernment interest in psi and the level of research funding of the field In WG Roll (Ed) Re-search in Parapsychology 1978 Metuchen NJ Scarecrow Press pp 54

Tart C T amp Labore C M (1986) Attitudes towards strongly functioning psi A preliminary sur-vey Journal of the America Society for Psychical Research 80 163

Tart C T (1986) Psychicsrsquo fears of psychic powers Journal of the American Society for Psychi-cal Research 80 278

Watt C A McDermott Z amp Ravenscroft J (1997) Exploring the limits of remote mental inter-action A comparison of blocking versus co-operating strategies In R Wiseman (Ed) Pro-ceedings of the 40th Annual Convention of the Parapsychological Association held in conjunc-tion with the Society for Psychical Research Brighton England pp 464ndash476

Blocking vs Co-operating 535

that the exploratory variance measure does not reveal any additional indica-tion of the operation of a remote mental influence in this study However thewider range of variance scoring continues to suggest that this measure mayhave greater potential sensitivity than the summary PSI score

Influenceesrsquo Blocking and Co-operating Strategies

We were interested in learning more about what particular strategies partici-pants chose to adopt when blocking and co-operating As in Study 1 a post-session questionnaire was used to gather this information Table 8 shows thepercentage of participants who adopted each kind of strategy

The table shows a very similar pattern to that found in Study 1 The mainpoint of interest is that almost no participants took a passive approach toblocking (ie adopting an openrelaxing strategy) A much higher proportionpreferred to adopt this approach when co-operating The preferred mode forboth conditions was to use some form of visual imagery such as imagining ashield or a screen when blocking and imagining an open channel such as ariver when co-operating

The majority of participants reported no concerns over the possibility ofhaving their physiology remotely influenced by another person as found instudy one

Sex Effects

As in study one the sexes were not equally represented among participantsWhereas Study 1 had an excess of male participants in Study 2 twice as manyfemales (N = 33) as males (N = 17) participated It was therefore important tolook for sex differences since any such differences could confound the inter-pretation of overall trends We found no difference between sexes when co-op-erating (female mean PSI score = 5008 male mean PSI score = 5006 t = 107 df = 48) There was a nonsignificant trend for females to have higher

Blocking vs Co-operating 529

TABLE 7Summary Data Mean Variance Scores(N = 50 MCE = 50 range in brackets)

Co-operate Variance Block Variance Combined Variance

5076 5054 5034(3164- 7279) (3572- 6813) (3749- 6671)

TABLE 8Influenceesrsquo Mental Strategies ()

Visual Imagery Cognitive VisCog OpenRelax

Blocking 64 26 6 2Co-operating 60 8 12 20

530 C Watt et al

PSI scores than males when blocking (mean PSI scores 5030 and 4998 re-spectively t = 992 df = 48) Combining blocking and co-operating conditionsthere was no overall difference between the sexes (mean PSI scores 5020 forfemales 5008 for males t = 665 df = 48)

Experimenter Effects

Study 2 found a statistically significant overall remote influence effect(combined PSI = 50155 t = 1806) Table 9 shows the breakdown of remoteinfluence scores for the two experimenters CW and JR Looking at the overalldata for the two experimenters we see that CWrsquos sessions were associatedwith most of the remote influence effect However due to the smaller N CWrsquossessions are not independently significant (t = 1632) and there is not a signifi-cant difference between CW and JRrsquos overall PSI scores (t = 678) Post hocwe calculated effect size r according to the formula r = [t2(t2+df)]12 for the twoexperimenters For JR (t = 817 df = 22) we found a small effect sizer = +17 For CW (t = 1632 df = 26) the effect size was moderate r = +30This may be some post hoc indication of an experimenter effect in Study 2

InfluenceeInfluencer Sex Pairings

As can be seen from Table 10 Study 2 found exactly the opposite trend tothat found in Study 1 On a post hoc test same sex pairings scored significant-ly higher on the remote influence PSI measure than mixed sex pairings(t = - 2034 df = 48 p = 048 2 - t) This contradicts Schlitz and LaBergersquos(1997) finding of greater autonomic detection of remote staring for mixed sexpairs

TABLE 9Remote Influence Results by Experimenter

Experimenter N sessions Coop PSI Block PSI Combined PSI

CW 27 5011 5026 5021JR 23 5003 5010 5009

TABLE 10Mean PSI Scores for Different InfluenceeInfluencer Sex Pairings

Pairing N Coop PSI Block PSI Combined PSI

MF 11 4977 5009 4997FM 17 4983 5021 5003FF 16 5035 5038 5038MM 6 5060 4978 5027

General Summary and Conclusions

We have described the findings of two studies exploring the limits of remotemental influence effects We wished to see whether the mental strategiesadopted by the influencee would affect remote influence scores In particularcould an influencee block or resist a remote influence attempt We recorded in-fluenceesrsquo electrodermal activity while the influencer attempted to calm to ac-tivate or to have no influence over the influenceesrsquo physiology In each studythere were two conditions reflecting the influenceesrsquo designated mental strate-gies in response to the remote influence attempt blocking and co-operating

There were only two formal hypotheses 1) that remote influence scoreswould be significantly higher during the co-operating condition than duringthe blocking condition and 2) that overall there would be significant evidenceof remote mental influence on physiology in the designated direction (iecalming when calming was the influencerrsquos intention activating when activat-ing was the influencerrsquos intention) Contrary to prediction Study 1 found thatremote influence effects were higher in the block condition than in the co-op-erate condition (but not to a significant extent) Also overall remote influenceperformance was at chance level This study therefore found no support for ei-ther of the formal hypotheses Study 2 consisted of a replication of Study 1with a few minor adjustments plus additional measures that were not reportedin this paper Again contrary to prediction remote influence scores were non-significantly higher in the block condition than in the co-operate conditionThis replicates the trend found in Study 1 failing to support hypothesis oneThese results suggest that the mental strategy adopted by the influencee in theface of a remote influence attempt has no effect on remote influence perfor-mance Referring back to the questions posed in the introduction then wehave not found any evidence to suggest that it is possible to resist a remote in-fluence effect and there is as yet no indication of the possible limiting condi-tions for remote mental influence

Unlike Study 1 Study 2 found statistically significant evidence of an overallremote influence effect in the prespecified direction (ie calming of EDAwhen that was the influencerrsquos intention activation of EDA when that was theinfluencerrsquos intention) In order to facilitate comparison with a recent meta-an-alytic review of remote influence studies we calculated the effect size (r) ac-cording to the formula r = [t2(t2+df)]12 For their review of 19 EDA influencestudies Schlitz and Braud (1997) found an average effect size of r = +25 OurStudy 1 with overall t = - 031 df = 31 gives a negligible effect size of r = - 006 Study 2 with overall t = 1806 df = 49 gives r = +25 Study 2 there-fore replicates the mean effect size found in EDA studies conducted by otherresearchers in other parapsychology laboratories Post hoc we can combinethe data of the two studies for greater statistical power Overall t = 1176 df = 81 giving a small effect size of r = +13

We can find no clear reason why Study 2 found an EDA influence effectwhile Study 1 did not A number of factors are worth considering 1) The

Blocking vs Co-operating 531

532 C Watt et al

significant result could be due to chance or undetected artifact 2) Study 2 im-plemented what we considered to be minor methodological improvements thatmight have removed some artifactual fluctuations from the EDA data that weobserved in study one 3) Each experimenter in Study 2 conducted three testingsessions with their participant By the third session in which the remote influ-ence measure was taken experimenters were on quite friendly terms with theirparticipants perhaps more so than in Study 1 where each participant came infor only one testing session 4) The participants in Study 2 were more represen-tative of the general population we felt than those in Study 1 Study 1 in-volved more undergraduate psychology students and associates of the experi-menters 5) The experimenters in study one were less experienced with theremote influence protocol than when they conducted Study 2

In addition to the formal hypotheses we reported a number of exploratorymeasures Study 1 found that influenceesrsquo self-reported vividness of visual im-agery appeared to be unrelated to the remote influence performance In bothstudies influenceesrsquo qualitative descriptions of the actual mental strategiesthey employed when blocking and co-operating revealed a tendency for differ-ent strategies to be preferred for the different conditions The use of visual im-agery was popular for both blocking (eg visualizing being surrounded by aprotective cocoon) and co-operating (eg visualizing a river flowing betweeninfluencer and influencee) However a considerable proportion of participantsreported adopting an open or relaxed frame of mind when co-operating whilstalmost no participants chose this strategy when blocking This was despite thefact that prior to each session we encouraged participants to adopt comparablestrategies for blocking and co-operating

It was interesting and perhaps surprising to note that in both studies themajority of participants reported no feelings of concern or discomfort at thepossibility of having their physiology remotely influenced Those who wereconcerned tended to be those who knew more about parapsychology and thepossible harmful applications of remote influence A small proportion of par-ticipants was unconcerned because they did not believe such an influence waspossible in the first place

Both studies employed an exploratory variance measure of EDA influenceThis measure showed a greater range of scoring overall thus perhaps greaterpotential sensitivity to a remote influence effect However the variance mea-sure showed no indication of a remote influence effect To date researchershave employed a variety of indicators of remote influence as the literaturegrows progress may be made on identifying which is the most appropriate in-dicator

Neither study found significant sex differences in remote influence effectsthough in both studies female influencees tended to have higher remote influ-ence scores than male influencees

In terms of their overall remote influence effects there appeared to be nodifference between experimenterinfluencers in Study 1 (r = +06 - 05 and

- 06) In Study 2 effect size measures indicated that the experimenters dif-fered in remote influence effect sizes (r = +17 and +30) Due to the smaller Nthis difference was not found to be statistically significant Post hoc combin-ing Studies 1 and 2 for experimenters JW and CW (ZMcD was not an experi-menter in Study 2) the effect sizes are r = +08 and r = +20 respectively

We conducted a post hoc analysis of the effects of influenceeinfluencer sexpairings on remote influence scores Our two studies came up with contradic-tory findings each statistically significant Study one found that mixed sexpairings had significantly higher remote influence scores than same sex pair-ings This replicated the post hoc finding of Schlitz and LaBerge (1997) in con-nection with the autonomic detection of remote observation Study 2 foundthat same sex pairings had significantly higher remote influence scores thanmixed sex pairings Our contradictory findings do not help to clarify the pic-ture on the importance or otherwise of sex pairings It is quite possible that sexis not the most salient aspect of the dynamic between pairs Age personalityand familiarity are likely to be important variables moderating the relationshipbetween influencer and influencee

There are many more variables that we could look at but as the number ofanalyses multiplies so does the opportunity for spurious significant resultsFor this reason we have named only two formal hypotheses for each experi-ment All other analyses are regarded as exploratory and any trends that wehave highlighted above would require replication

In conclusion these two studies have found no evidence of a difference inremote influence effects when influencees are asked to adopt cooperative men-tal strategies compared to when they are asked to adopt blocking mental strate-gies In fact contrary to expectation both studies found slightly higher remoteinfluence effects for the blocking condition compared to the co-operating con-dition On this basis we have found no indication of limiting conditions for re-mote influence effects Therefore while anecdotal advice exists on how to de-fend oneself against a remote physiological influence attempt in the realworld laboratory evidence to date has provided no support for this advice

Given the importance of this question both in practical terms and in theoret-ical terms we argue that this line of research deserves continued investigationIf a way could be found to conduct research with a more realistic scenario mdash where the remote influence was not just neutrally and abstractlydefined as unwelcome mdash then our research might have more meaningful thingsto say about the parameters of remote influence It is interesting and perhapssurprising to note that when Watt has lectured on this topic to informed lay au-diences the feedback has often been that researchers must use more potential-ly harmful or disruptive remote influence (of course with the informed consentof the participants) While there are clear ethical inhibitions when consideringpossible physiological influence perhaps research using higher level influencewould be more acceptable For example one could attempt to disrupt or en-hance the concentration of chess players who were aware of the influence

Blocking vs Co-operating 533

534 C Watt et al

attempt and who were concurrently blocking or co-operating with the influ-ence attempt Research already exists to suggest that remote influence does notjust operate at physiological levels Attention focusing for example can be fa-cilitated by a remote helper (Brady amp Morris 1997 Braud et al 1995) Futureresearch in this area must therefore expand beyond physiological influences

Acknowledgements

The authors gratefully acknowledge the financial support they receivedfrom the Institut fuumlr Grenzgebiete der Psychologie und PsychohygieneFreiburg Germany while conducting these studies We would also like tothank Dr Paul Stevens for generously helping to solve software difficultieswhen preparing for these studies and the anonymous reviewer for helpfulcomments Address reprint requests to Dr C Watt Department of PsychologyUniversity of Edinburgh 7 George Square Edinburgh EH8 9JZ Scotland UK

ReferencesBrady C amp Morris R L (1997) Attention focusing facilitated through remote mental interac-

tion A replication and exploration of parameters In R Wiseman (Ed) Proceedings of Pre-sented Papers The Parapsychological Association 40th Annual Convention held in conjunc-tion with The Society for Psychical Research pp 73ndash91

Braud W G (1985) Blockingshielding psychic functioning through psychological and psychictechniques A report of three preliminary studies In RA White and J Solfvin (Eds) Re-search in Parapsychology 1984 Metuchen NJ Scarecrow Press pp 42ndash44

Braud W G amp Schlitz M J (1991) Consciousness interactions with remote biological systemsAnomalous intentionality effects Subtle Energies 2 1

Braud W G Schlitz M J Collins J and Klitch H (1985) Further studies of the bio-PK effectFeedback blocking specificitygenerality In RA White amp J Solfvin (Eds) Research inParapsychology 1984 Metuchen NJ Scarecrow Press pp 45ndash48

Braud W G Shafer D McNeill K and Guerra V (1995) Attention focusing facilitatedthrough remote mental interaction Journal of the American Society for Psychical Research89 103

Cohen J (1997) Statistical Power Analysis for the Behavioral Sciences New York AcademicPress

Dalton K (1994) Remotely influenced ESP performance in a computer task A preliminarystudy Proceedings of the 37th Annual Convention of the Parapsychological Association pp95ndash103

Dossey L (1994) Healing and the mind Is there a dark side Journal of Scientific Exploration 873

Halifax-Grof J (1974) Hex death In A Angoff amp D Barth (Eds) Parapsychology and Anthro-pology Proceedings of an International Conference Held in London England August 29ndash311973 NY Parapsychology Foundation

Hope M (1983) Practical Techniques of Psychic Self-Defence Wellingborough Northampton-shire Aquarian Press

Marks DF (1973) Visual imagery differences in the recall of pictures British Journal of Psy-chology 64 17

Nash CB (1982) Psychokinetic control of bacterial growth Journal of the Society for PsychicalResearch 51 217

Radin DI Taylor RK and Braud WG (1995) Remote mental influence of human electroder-mal activity A pilot replication European Journal of Parapsychology 11 19

Schlitz M amp Braud W G (1997) Distant intentionality and healing Assessing the evidence Al-ternative Therapies 3 62

Schlitz M J amp LaBerge S (1997) Covert observation increases skin conductance in subjectsunaware of when they are being observed A replication Journal of Parapsychology 61 185

Schmeidler G R (1958) Agent-percipient relationships Journal of the American Society forPsychical Research 52 47

Schmeidler G R (1961) Evidence for two kinds of telepathy International Journal of Parapsy-chology 3 5

Tart C T (1979) A survey of expert opinion on potentially negative uses of psi United Statesgovernment interest in psi and the level of research funding of the field In WG Roll (Ed) Re-search in Parapsychology 1978 Metuchen NJ Scarecrow Press pp 54

Tart C T amp Labore C M (1986) Attitudes towards strongly functioning psi A preliminary sur-vey Journal of the America Society for Psychical Research 80 163

Tart C T (1986) Psychicsrsquo fears of psychic powers Journal of the American Society for Psychi-cal Research 80 278

Watt C A McDermott Z amp Ravenscroft J (1997) Exploring the limits of remote mental inter-action A comparison of blocking versus co-operating strategies In R Wiseman (Ed) Pro-ceedings of the 40th Annual Convention of the Parapsychological Association held in conjunc-tion with the Society for Psychical Research Brighton England pp 464ndash476

Blocking vs Co-operating 535

530 C Watt et al

PSI scores than males when blocking (mean PSI scores 5030 and 4998 re-spectively t = 992 df = 48) Combining blocking and co-operating conditionsthere was no overall difference between the sexes (mean PSI scores 5020 forfemales 5008 for males t = 665 df = 48)

Experimenter Effects

Study 2 found a statistically significant overall remote influence effect(combined PSI = 50155 t = 1806) Table 9 shows the breakdown of remoteinfluence scores for the two experimenters CW and JR Looking at the overalldata for the two experimenters we see that CWrsquos sessions were associatedwith most of the remote influence effect However due to the smaller N CWrsquossessions are not independently significant (t = 1632) and there is not a signifi-cant difference between CW and JRrsquos overall PSI scores (t = 678) Post hocwe calculated effect size r according to the formula r = [t2(t2+df)]12 for the twoexperimenters For JR (t = 817 df = 22) we found a small effect sizer = +17 For CW (t = 1632 df = 26) the effect size was moderate r = +30This may be some post hoc indication of an experimenter effect in Study 2

InfluenceeInfluencer Sex Pairings

As can be seen from Table 10 Study 2 found exactly the opposite trend tothat found in Study 1 On a post hoc test same sex pairings scored significant-ly higher on the remote influence PSI measure than mixed sex pairings(t = - 2034 df = 48 p = 048 2 - t) This contradicts Schlitz and LaBergersquos(1997) finding of greater autonomic detection of remote staring for mixed sexpairs

TABLE 9Remote Influence Results by Experimenter

Experimenter N sessions Coop PSI Block PSI Combined PSI

CW 27 5011 5026 5021JR 23 5003 5010 5009

TABLE 10Mean PSI Scores for Different InfluenceeInfluencer Sex Pairings

Pairing N Coop PSI Block PSI Combined PSI

MF 11 4977 5009 4997FM 17 4983 5021 5003FF 16 5035 5038 5038MM 6 5060 4978 5027

General Summary and Conclusions

We have described the findings of two studies exploring the limits of remotemental influence effects We wished to see whether the mental strategiesadopted by the influencee would affect remote influence scores In particularcould an influencee block or resist a remote influence attempt We recorded in-fluenceesrsquo electrodermal activity while the influencer attempted to calm to ac-tivate or to have no influence over the influenceesrsquo physiology In each studythere were two conditions reflecting the influenceesrsquo designated mental strate-gies in response to the remote influence attempt blocking and co-operating

There were only two formal hypotheses 1) that remote influence scoreswould be significantly higher during the co-operating condition than duringthe blocking condition and 2) that overall there would be significant evidenceof remote mental influence on physiology in the designated direction (iecalming when calming was the influencerrsquos intention activating when activat-ing was the influencerrsquos intention) Contrary to prediction Study 1 found thatremote influence effects were higher in the block condition than in the co-op-erate condition (but not to a significant extent) Also overall remote influenceperformance was at chance level This study therefore found no support for ei-ther of the formal hypotheses Study 2 consisted of a replication of Study 1with a few minor adjustments plus additional measures that were not reportedin this paper Again contrary to prediction remote influence scores were non-significantly higher in the block condition than in the co-operate conditionThis replicates the trend found in Study 1 failing to support hypothesis oneThese results suggest that the mental strategy adopted by the influencee in theface of a remote influence attempt has no effect on remote influence perfor-mance Referring back to the questions posed in the introduction then wehave not found any evidence to suggest that it is possible to resist a remote in-fluence effect and there is as yet no indication of the possible limiting condi-tions for remote mental influence

Unlike Study 1 Study 2 found statistically significant evidence of an overallremote influence effect in the prespecified direction (ie calming of EDAwhen that was the influencerrsquos intention activation of EDA when that was theinfluencerrsquos intention) In order to facilitate comparison with a recent meta-an-alytic review of remote influence studies we calculated the effect size (r) ac-cording to the formula r = [t2(t2+df)]12 For their review of 19 EDA influencestudies Schlitz and Braud (1997) found an average effect size of r = +25 OurStudy 1 with overall t = - 031 df = 31 gives a negligible effect size of r = - 006 Study 2 with overall t = 1806 df = 49 gives r = +25 Study 2 there-fore replicates the mean effect size found in EDA studies conducted by otherresearchers in other parapsychology laboratories Post hoc we can combinethe data of the two studies for greater statistical power Overall t = 1176 df = 81 giving a small effect size of r = +13

We can find no clear reason why Study 2 found an EDA influence effectwhile Study 1 did not A number of factors are worth considering 1) The

Blocking vs Co-operating 531

532 C Watt et al

significant result could be due to chance or undetected artifact 2) Study 2 im-plemented what we considered to be minor methodological improvements thatmight have removed some artifactual fluctuations from the EDA data that weobserved in study one 3) Each experimenter in Study 2 conducted three testingsessions with their participant By the third session in which the remote influ-ence measure was taken experimenters were on quite friendly terms with theirparticipants perhaps more so than in Study 1 where each participant came infor only one testing session 4) The participants in Study 2 were more represen-tative of the general population we felt than those in Study 1 Study 1 in-volved more undergraduate psychology students and associates of the experi-menters 5) The experimenters in study one were less experienced with theremote influence protocol than when they conducted Study 2

In addition to the formal hypotheses we reported a number of exploratorymeasures Study 1 found that influenceesrsquo self-reported vividness of visual im-agery appeared to be unrelated to the remote influence performance In bothstudies influenceesrsquo qualitative descriptions of the actual mental strategiesthey employed when blocking and co-operating revealed a tendency for differ-ent strategies to be preferred for the different conditions The use of visual im-agery was popular for both blocking (eg visualizing being surrounded by aprotective cocoon) and co-operating (eg visualizing a river flowing betweeninfluencer and influencee) However a considerable proportion of participantsreported adopting an open or relaxed frame of mind when co-operating whilstalmost no participants chose this strategy when blocking This was despite thefact that prior to each session we encouraged participants to adopt comparablestrategies for blocking and co-operating

It was interesting and perhaps surprising to note that in both studies themajority of participants reported no feelings of concern or discomfort at thepossibility of having their physiology remotely influenced Those who wereconcerned tended to be those who knew more about parapsychology and thepossible harmful applications of remote influence A small proportion of par-ticipants was unconcerned because they did not believe such an influence waspossible in the first place

Both studies employed an exploratory variance measure of EDA influenceThis measure showed a greater range of scoring overall thus perhaps greaterpotential sensitivity to a remote influence effect However the variance mea-sure showed no indication of a remote influence effect To date researchershave employed a variety of indicators of remote influence as the literaturegrows progress may be made on identifying which is the most appropriate in-dicator

Neither study found significant sex differences in remote influence effectsthough in both studies female influencees tended to have higher remote influ-ence scores than male influencees

In terms of their overall remote influence effects there appeared to be nodifference between experimenterinfluencers in Study 1 (r = +06 - 05 and

- 06) In Study 2 effect size measures indicated that the experimenters dif-fered in remote influence effect sizes (r = +17 and +30) Due to the smaller Nthis difference was not found to be statistically significant Post hoc combin-ing Studies 1 and 2 for experimenters JW and CW (ZMcD was not an experi-menter in Study 2) the effect sizes are r = +08 and r = +20 respectively

We conducted a post hoc analysis of the effects of influenceeinfluencer sexpairings on remote influence scores Our two studies came up with contradic-tory findings each statistically significant Study one found that mixed sexpairings had significantly higher remote influence scores than same sex pair-ings This replicated the post hoc finding of Schlitz and LaBerge (1997) in con-nection with the autonomic detection of remote observation Study 2 foundthat same sex pairings had significantly higher remote influence scores thanmixed sex pairings Our contradictory findings do not help to clarify the pic-ture on the importance or otherwise of sex pairings It is quite possible that sexis not the most salient aspect of the dynamic between pairs Age personalityand familiarity are likely to be important variables moderating the relationshipbetween influencer and influencee

There are many more variables that we could look at but as the number ofanalyses multiplies so does the opportunity for spurious significant resultsFor this reason we have named only two formal hypotheses for each experi-ment All other analyses are regarded as exploratory and any trends that wehave highlighted above would require replication

In conclusion these two studies have found no evidence of a difference inremote influence effects when influencees are asked to adopt cooperative men-tal strategies compared to when they are asked to adopt blocking mental strate-gies In fact contrary to expectation both studies found slightly higher remoteinfluence effects for the blocking condition compared to the co-operating con-dition On this basis we have found no indication of limiting conditions for re-mote influence effects Therefore while anecdotal advice exists on how to de-fend oneself against a remote physiological influence attempt in the realworld laboratory evidence to date has provided no support for this advice

Given the importance of this question both in practical terms and in theoret-ical terms we argue that this line of research deserves continued investigationIf a way could be found to conduct research with a more realistic scenario mdash where the remote influence was not just neutrally and abstractlydefined as unwelcome mdash then our research might have more meaningful thingsto say about the parameters of remote influence It is interesting and perhapssurprising to note that when Watt has lectured on this topic to informed lay au-diences the feedback has often been that researchers must use more potential-ly harmful or disruptive remote influence (of course with the informed consentof the participants) While there are clear ethical inhibitions when consideringpossible physiological influence perhaps research using higher level influencewould be more acceptable For example one could attempt to disrupt or en-hance the concentration of chess players who were aware of the influence

Blocking vs Co-operating 533

534 C Watt et al

attempt and who were concurrently blocking or co-operating with the influ-ence attempt Research already exists to suggest that remote influence does notjust operate at physiological levels Attention focusing for example can be fa-cilitated by a remote helper (Brady amp Morris 1997 Braud et al 1995) Futureresearch in this area must therefore expand beyond physiological influences

Acknowledgements

The authors gratefully acknowledge the financial support they receivedfrom the Institut fuumlr Grenzgebiete der Psychologie und PsychohygieneFreiburg Germany while conducting these studies We would also like tothank Dr Paul Stevens for generously helping to solve software difficultieswhen preparing for these studies and the anonymous reviewer for helpfulcomments Address reprint requests to Dr C Watt Department of PsychologyUniversity of Edinburgh 7 George Square Edinburgh EH8 9JZ Scotland UK

ReferencesBrady C amp Morris R L (1997) Attention focusing facilitated through remote mental interac-

tion A replication and exploration of parameters In R Wiseman (Ed) Proceedings of Pre-sented Papers The Parapsychological Association 40th Annual Convention held in conjunc-tion with The Society for Psychical Research pp 73ndash91

Braud W G (1985) Blockingshielding psychic functioning through psychological and psychictechniques A report of three preliminary studies In RA White and J Solfvin (Eds) Re-search in Parapsychology 1984 Metuchen NJ Scarecrow Press pp 42ndash44

Braud W G amp Schlitz M J (1991) Consciousness interactions with remote biological systemsAnomalous intentionality effects Subtle Energies 2 1

Braud W G Schlitz M J Collins J and Klitch H (1985) Further studies of the bio-PK effectFeedback blocking specificitygenerality In RA White amp J Solfvin (Eds) Research inParapsychology 1984 Metuchen NJ Scarecrow Press pp 45ndash48

Braud W G Shafer D McNeill K and Guerra V (1995) Attention focusing facilitatedthrough remote mental interaction Journal of the American Society for Psychical Research89 103

Cohen J (1997) Statistical Power Analysis for the Behavioral Sciences New York AcademicPress

Dalton K (1994) Remotely influenced ESP performance in a computer task A preliminarystudy Proceedings of the 37th Annual Convention of the Parapsychological Association pp95ndash103

Dossey L (1994) Healing and the mind Is there a dark side Journal of Scientific Exploration 873

Halifax-Grof J (1974) Hex death In A Angoff amp D Barth (Eds) Parapsychology and Anthro-pology Proceedings of an International Conference Held in London England August 29ndash311973 NY Parapsychology Foundation

Hope M (1983) Practical Techniques of Psychic Self-Defence Wellingborough Northampton-shire Aquarian Press

Marks DF (1973) Visual imagery differences in the recall of pictures British Journal of Psy-chology 64 17

Nash CB (1982) Psychokinetic control of bacterial growth Journal of the Society for PsychicalResearch 51 217

Radin DI Taylor RK and Braud WG (1995) Remote mental influence of human electroder-mal activity A pilot replication European Journal of Parapsychology 11 19

Schlitz M amp Braud W G (1997) Distant intentionality and healing Assessing the evidence Al-ternative Therapies 3 62

Schlitz M J amp LaBerge S (1997) Covert observation increases skin conductance in subjectsunaware of when they are being observed A replication Journal of Parapsychology 61 185

Schmeidler G R (1958) Agent-percipient relationships Journal of the American Society forPsychical Research 52 47

Schmeidler G R (1961) Evidence for two kinds of telepathy International Journal of Parapsy-chology 3 5

Tart C T (1979) A survey of expert opinion on potentially negative uses of psi United Statesgovernment interest in psi and the level of research funding of the field In WG Roll (Ed) Re-search in Parapsychology 1978 Metuchen NJ Scarecrow Press pp 54

Tart C T amp Labore C M (1986) Attitudes towards strongly functioning psi A preliminary sur-vey Journal of the America Society for Psychical Research 80 163

Tart C T (1986) Psychicsrsquo fears of psychic powers Journal of the American Society for Psychi-cal Research 80 278

Watt C A McDermott Z amp Ravenscroft J (1997) Exploring the limits of remote mental inter-action A comparison of blocking versus co-operating strategies In R Wiseman (Ed) Pro-ceedings of the 40th Annual Convention of the Parapsychological Association held in conjunc-tion with the Society for Psychical Research Brighton England pp 464ndash476

Blocking vs Co-operating 535

General Summary and Conclusions

We have described the findings of two studies exploring the limits of remotemental influence effects We wished to see whether the mental strategiesadopted by the influencee would affect remote influence scores In particularcould an influencee block or resist a remote influence attempt We recorded in-fluenceesrsquo electrodermal activity while the influencer attempted to calm to ac-tivate or to have no influence over the influenceesrsquo physiology In each studythere were two conditions reflecting the influenceesrsquo designated mental strate-gies in response to the remote influence attempt blocking and co-operating

There were only two formal hypotheses 1) that remote influence scoreswould be significantly higher during the co-operating condition than duringthe blocking condition and 2) that overall there would be significant evidenceof remote mental influence on physiology in the designated direction (iecalming when calming was the influencerrsquos intention activating when activat-ing was the influencerrsquos intention) Contrary to prediction Study 1 found thatremote influence effects were higher in the block condition than in the co-op-erate condition (but not to a significant extent) Also overall remote influenceperformance was at chance level This study therefore found no support for ei-ther of the formal hypotheses Study 2 consisted of a replication of Study 1with a few minor adjustments plus additional measures that were not reportedin this paper Again contrary to prediction remote influence scores were non-significantly higher in the block condition than in the co-operate conditionThis replicates the trend found in Study 1 failing to support hypothesis oneThese results suggest that the mental strategy adopted by the influencee in theface of a remote influence attempt has no effect on remote influence perfor-mance Referring back to the questions posed in the introduction then wehave not found any evidence to suggest that it is possible to resist a remote in-fluence effect and there is as yet no indication of the possible limiting condi-tions for remote mental influence

Unlike Study 1 Study 2 found statistically significant evidence of an overallremote influence effect in the prespecified direction (ie calming of EDAwhen that was the influencerrsquos intention activation of EDA when that was theinfluencerrsquos intention) In order to facilitate comparison with a recent meta-an-alytic review of remote influence studies we calculated the effect size (r) ac-cording to the formula r = [t2(t2+df)]12 For their review of 19 EDA influencestudies Schlitz and Braud (1997) found an average effect size of r = +25 OurStudy 1 with overall t = - 031 df = 31 gives a negligible effect size of r = - 006 Study 2 with overall t = 1806 df = 49 gives r = +25 Study 2 there-fore replicates the mean effect size found in EDA studies conducted by otherresearchers in other parapsychology laboratories Post hoc we can combinethe data of the two studies for greater statistical power Overall t = 1176 df = 81 giving a small effect size of r = +13

We can find no clear reason why Study 2 found an EDA influence effectwhile Study 1 did not A number of factors are worth considering 1) The

Blocking vs Co-operating 531

532 C Watt et al

significant result could be due to chance or undetected artifact 2) Study 2 im-plemented what we considered to be minor methodological improvements thatmight have removed some artifactual fluctuations from the EDA data that weobserved in study one 3) Each experimenter in Study 2 conducted three testingsessions with their participant By the third session in which the remote influ-ence measure was taken experimenters were on quite friendly terms with theirparticipants perhaps more so than in Study 1 where each participant came infor only one testing session 4) The participants in Study 2 were more represen-tative of the general population we felt than those in Study 1 Study 1 in-volved more undergraduate psychology students and associates of the experi-menters 5) The experimenters in study one were less experienced with theremote influence protocol than when they conducted Study 2

In addition to the formal hypotheses we reported a number of exploratorymeasures Study 1 found that influenceesrsquo self-reported vividness of visual im-agery appeared to be unrelated to the remote influence performance In bothstudies influenceesrsquo qualitative descriptions of the actual mental strategiesthey employed when blocking and co-operating revealed a tendency for differ-ent strategies to be preferred for the different conditions The use of visual im-agery was popular for both blocking (eg visualizing being surrounded by aprotective cocoon) and co-operating (eg visualizing a river flowing betweeninfluencer and influencee) However a considerable proportion of participantsreported adopting an open or relaxed frame of mind when co-operating whilstalmost no participants chose this strategy when blocking This was despite thefact that prior to each session we encouraged participants to adopt comparablestrategies for blocking and co-operating

It was interesting and perhaps surprising to note that in both studies themajority of participants reported no feelings of concern or discomfort at thepossibility of having their physiology remotely influenced Those who wereconcerned tended to be those who knew more about parapsychology and thepossible harmful applications of remote influence A small proportion of par-ticipants was unconcerned because they did not believe such an influence waspossible in the first place

Both studies employed an exploratory variance measure of EDA influenceThis measure showed a greater range of scoring overall thus perhaps greaterpotential sensitivity to a remote influence effect However the variance mea-sure showed no indication of a remote influence effect To date researchershave employed a variety of indicators of remote influence as the literaturegrows progress may be made on identifying which is the most appropriate in-dicator

Neither study found significant sex differences in remote influence effectsthough in both studies female influencees tended to have higher remote influ-ence scores than male influencees

In terms of their overall remote influence effects there appeared to be nodifference between experimenterinfluencers in Study 1 (r = +06 - 05 and

- 06) In Study 2 effect size measures indicated that the experimenters dif-fered in remote influence effect sizes (r = +17 and +30) Due to the smaller Nthis difference was not found to be statistically significant Post hoc combin-ing Studies 1 and 2 for experimenters JW and CW (ZMcD was not an experi-menter in Study 2) the effect sizes are r = +08 and r = +20 respectively

We conducted a post hoc analysis of the effects of influenceeinfluencer sexpairings on remote influence scores Our two studies came up with contradic-tory findings each statistically significant Study one found that mixed sexpairings had significantly higher remote influence scores than same sex pair-ings This replicated the post hoc finding of Schlitz and LaBerge (1997) in con-nection with the autonomic detection of remote observation Study 2 foundthat same sex pairings had significantly higher remote influence scores thanmixed sex pairings Our contradictory findings do not help to clarify the pic-ture on the importance or otherwise of sex pairings It is quite possible that sexis not the most salient aspect of the dynamic between pairs Age personalityand familiarity are likely to be important variables moderating the relationshipbetween influencer and influencee

There are many more variables that we could look at but as the number ofanalyses multiplies so does the opportunity for spurious significant resultsFor this reason we have named only two formal hypotheses for each experi-ment All other analyses are regarded as exploratory and any trends that wehave highlighted above would require replication

In conclusion these two studies have found no evidence of a difference inremote influence effects when influencees are asked to adopt cooperative men-tal strategies compared to when they are asked to adopt blocking mental strate-gies In fact contrary to expectation both studies found slightly higher remoteinfluence effects for the blocking condition compared to the co-operating con-dition On this basis we have found no indication of limiting conditions for re-mote influence effects Therefore while anecdotal advice exists on how to de-fend oneself against a remote physiological influence attempt in the realworld laboratory evidence to date has provided no support for this advice

Given the importance of this question both in practical terms and in theoret-ical terms we argue that this line of research deserves continued investigationIf a way could be found to conduct research with a more realistic scenario mdash where the remote influence was not just neutrally and abstractlydefined as unwelcome mdash then our research might have more meaningful thingsto say about the parameters of remote influence It is interesting and perhapssurprising to note that when Watt has lectured on this topic to informed lay au-diences the feedback has often been that researchers must use more potential-ly harmful or disruptive remote influence (of course with the informed consentof the participants) While there are clear ethical inhibitions when consideringpossible physiological influence perhaps research using higher level influencewould be more acceptable For example one could attempt to disrupt or en-hance the concentration of chess players who were aware of the influence

Blocking vs Co-operating 533

534 C Watt et al

attempt and who were concurrently blocking or co-operating with the influ-ence attempt Research already exists to suggest that remote influence does notjust operate at physiological levels Attention focusing for example can be fa-cilitated by a remote helper (Brady amp Morris 1997 Braud et al 1995) Futureresearch in this area must therefore expand beyond physiological influences

Acknowledgements

The authors gratefully acknowledge the financial support they receivedfrom the Institut fuumlr Grenzgebiete der Psychologie und PsychohygieneFreiburg Germany while conducting these studies We would also like tothank Dr Paul Stevens for generously helping to solve software difficultieswhen preparing for these studies and the anonymous reviewer for helpfulcomments Address reprint requests to Dr C Watt Department of PsychologyUniversity of Edinburgh 7 George Square Edinburgh EH8 9JZ Scotland UK

ReferencesBrady C amp Morris R L (1997) Attention focusing facilitated through remote mental interac-

tion A replication and exploration of parameters In R Wiseman (Ed) Proceedings of Pre-sented Papers The Parapsychological Association 40th Annual Convention held in conjunc-tion with The Society for Psychical Research pp 73ndash91

Braud W G (1985) Blockingshielding psychic functioning through psychological and psychictechniques A report of three preliminary studies In RA White and J Solfvin (Eds) Re-search in Parapsychology 1984 Metuchen NJ Scarecrow Press pp 42ndash44

Braud W G amp Schlitz M J (1991) Consciousness interactions with remote biological systemsAnomalous intentionality effects Subtle Energies 2 1

Braud W G Schlitz M J Collins J and Klitch H (1985) Further studies of the bio-PK effectFeedback blocking specificitygenerality In RA White amp J Solfvin (Eds) Research inParapsychology 1984 Metuchen NJ Scarecrow Press pp 45ndash48

Braud W G Shafer D McNeill K and Guerra V (1995) Attention focusing facilitatedthrough remote mental interaction Journal of the American Society for Psychical Research89 103

Cohen J (1997) Statistical Power Analysis for the Behavioral Sciences New York AcademicPress

Dalton K (1994) Remotely influenced ESP performance in a computer task A preliminarystudy Proceedings of the 37th Annual Convention of the Parapsychological Association pp95ndash103

Dossey L (1994) Healing and the mind Is there a dark side Journal of Scientific Exploration 873

Halifax-Grof J (1974) Hex death In A Angoff amp D Barth (Eds) Parapsychology and Anthro-pology Proceedings of an International Conference Held in London England August 29ndash311973 NY Parapsychology Foundation

Hope M (1983) Practical Techniques of Psychic Self-Defence Wellingborough Northampton-shire Aquarian Press

Marks DF (1973) Visual imagery differences in the recall of pictures British Journal of Psy-chology 64 17

Nash CB (1982) Psychokinetic control of bacterial growth Journal of the Society for PsychicalResearch 51 217

Radin DI Taylor RK and Braud WG (1995) Remote mental influence of human electroder-mal activity A pilot replication European Journal of Parapsychology 11 19

Schlitz M amp Braud W G (1997) Distant intentionality and healing Assessing the evidence Al-ternative Therapies 3 62

Schlitz M J amp LaBerge S (1997) Covert observation increases skin conductance in subjectsunaware of when they are being observed A replication Journal of Parapsychology 61 185

Schmeidler G R (1958) Agent-percipient relationships Journal of the American Society forPsychical Research 52 47

Schmeidler G R (1961) Evidence for two kinds of telepathy International Journal of Parapsy-chology 3 5

Tart C T (1979) A survey of expert opinion on potentially negative uses of psi United Statesgovernment interest in psi and the level of research funding of the field In WG Roll (Ed) Re-search in Parapsychology 1978 Metuchen NJ Scarecrow Press pp 54

Tart C T amp Labore C M (1986) Attitudes towards strongly functioning psi A preliminary sur-vey Journal of the America Society for Psychical Research 80 163

Tart C T (1986) Psychicsrsquo fears of psychic powers Journal of the American Society for Psychi-cal Research 80 278

Watt C A McDermott Z amp Ravenscroft J (1997) Exploring the limits of remote mental inter-action A comparison of blocking versus co-operating strategies In R Wiseman (Ed) Pro-ceedings of the 40th Annual Convention of the Parapsychological Association held in conjunc-tion with the Society for Psychical Research Brighton England pp 464ndash476

Blocking vs Co-operating 535

532 C Watt et al

significant result could be due to chance or undetected artifact 2) Study 2 im-plemented what we considered to be minor methodological improvements thatmight have removed some artifactual fluctuations from the EDA data that weobserved in study one 3) Each experimenter in Study 2 conducted three testingsessions with their participant By the third session in which the remote influ-ence measure was taken experimenters were on quite friendly terms with theirparticipants perhaps more so than in Study 1 where each participant came infor only one testing session 4) The participants in Study 2 were more represen-tative of the general population we felt than those in Study 1 Study 1 in-volved more undergraduate psychology students and associates of the experi-menters 5) The experimenters in study one were less experienced with theremote influence protocol than when they conducted Study 2

In addition to the formal hypotheses we reported a number of exploratorymeasures Study 1 found that influenceesrsquo self-reported vividness of visual im-agery appeared to be unrelated to the remote influence performance In bothstudies influenceesrsquo qualitative descriptions of the actual mental strategiesthey employed when blocking and co-operating revealed a tendency for differ-ent strategies to be preferred for the different conditions The use of visual im-agery was popular for both blocking (eg visualizing being surrounded by aprotective cocoon) and co-operating (eg visualizing a river flowing betweeninfluencer and influencee) However a considerable proportion of participantsreported adopting an open or relaxed frame of mind when co-operating whilstalmost no participants chose this strategy when blocking This was despite thefact that prior to each session we encouraged participants to adopt comparablestrategies for blocking and co-operating

It was interesting and perhaps surprising to note that in both studies themajority of participants reported no feelings of concern or discomfort at thepossibility of having their physiology remotely influenced Those who wereconcerned tended to be those who knew more about parapsychology and thepossible harmful applications of remote influence A small proportion of par-ticipants was unconcerned because they did not believe such an influence waspossible in the first place

Both studies employed an exploratory variance measure of EDA influenceThis measure showed a greater range of scoring overall thus perhaps greaterpotential sensitivity to a remote influence effect However the variance mea-sure showed no indication of a remote influence effect To date researchershave employed a variety of indicators of remote influence as the literaturegrows progress may be made on identifying which is the most appropriate in-dicator

Neither study found significant sex differences in remote influence effectsthough in both studies female influencees tended to have higher remote influ-ence scores than male influencees

In terms of their overall remote influence effects there appeared to be nodifference between experimenterinfluencers in Study 1 (r = +06 - 05 and

- 06) In Study 2 effect size measures indicated that the experimenters dif-fered in remote influence effect sizes (r = +17 and +30) Due to the smaller Nthis difference was not found to be statistically significant Post hoc combin-ing Studies 1 and 2 for experimenters JW and CW (ZMcD was not an experi-menter in Study 2) the effect sizes are r = +08 and r = +20 respectively

We conducted a post hoc analysis of the effects of influenceeinfluencer sexpairings on remote influence scores Our two studies came up with contradic-tory findings each statistically significant Study one found that mixed sexpairings had significantly higher remote influence scores than same sex pair-ings This replicated the post hoc finding of Schlitz and LaBerge (1997) in con-nection with the autonomic detection of remote observation Study 2 foundthat same sex pairings had significantly higher remote influence scores thanmixed sex pairings Our contradictory findings do not help to clarify the pic-ture on the importance or otherwise of sex pairings It is quite possible that sexis not the most salient aspect of the dynamic between pairs Age personalityand familiarity are likely to be important variables moderating the relationshipbetween influencer and influencee

There are many more variables that we could look at but as the number ofanalyses multiplies so does the opportunity for spurious significant resultsFor this reason we have named only two formal hypotheses for each experi-ment All other analyses are regarded as exploratory and any trends that wehave highlighted above would require replication

In conclusion these two studies have found no evidence of a difference inremote influence effects when influencees are asked to adopt cooperative men-tal strategies compared to when they are asked to adopt blocking mental strate-gies In fact contrary to expectation both studies found slightly higher remoteinfluence effects for the blocking condition compared to the co-operating con-dition On this basis we have found no indication of limiting conditions for re-mote influence effects Therefore while anecdotal advice exists on how to de-fend oneself against a remote physiological influence attempt in the realworld laboratory evidence to date has provided no support for this advice

Given the importance of this question both in practical terms and in theoret-ical terms we argue that this line of research deserves continued investigationIf a way could be found to conduct research with a more realistic scenario mdash where the remote influence was not just neutrally and abstractlydefined as unwelcome mdash then our research might have more meaningful thingsto say about the parameters of remote influence It is interesting and perhapssurprising to note that when Watt has lectured on this topic to informed lay au-diences the feedback has often been that researchers must use more potential-ly harmful or disruptive remote influence (of course with the informed consentof the participants) While there are clear ethical inhibitions when consideringpossible physiological influence perhaps research using higher level influencewould be more acceptable For example one could attempt to disrupt or en-hance the concentration of chess players who were aware of the influence

Blocking vs Co-operating 533

534 C Watt et al

attempt and who were concurrently blocking or co-operating with the influ-ence attempt Research already exists to suggest that remote influence does notjust operate at physiological levels Attention focusing for example can be fa-cilitated by a remote helper (Brady amp Morris 1997 Braud et al 1995) Futureresearch in this area must therefore expand beyond physiological influences

Acknowledgements

The authors gratefully acknowledge the financial support they receivedfrom the Institut fuumlr Grenzgebiete der Psychologie und PsychohygieneFreiburg Germany while conducting these studies We would also like tothank Dr Paul Stevens for generously helping to solve software difficultieswhen preparing for these studies and the anonymous reviewer for helpfulcomments Address reprint requests to Dr C Watt Department of PsychologyUniversity of Edinburgh 7 George Square Edinburgh EH8 9JZ Scotland UK

ReferencesBrady C amp Morris R L (1997) Attention focusing facilitated through remote mental interac-

tion A replication and exploration of parameters In R Wiseman (Ed) Proceedings of Pre-sented Papers The Parapsychological Association 40th Annual Convention held in conjunc-tion with The Society for Psychical Research pp 73ndash91

Braud W G (1985) Blockingshielding psychic functioning through psychological and psychictechniques A report of three preliminary studies In RA White and J Solfvin (Eds) Re-search in Parapsychology 1984 Metuchen NJ Scarecrow Press pp 42ndash44

Braud W G amp Schlitz M J (1991) Consciousness interactions with remote biological systemsAnomalous intentionality effects Subtle Energies 2 1

Braud W G Schlitz M J Collins J and Klitch H (1985) Further studies of the bio-PK effectFeedback blocking specificitygenerality In RA White amp J Solfvin (Eds) Research inParapsychology 1984 Metuchen NJ Scarecrow Press pp 45ndash48

Braud W G Shafer D McNeill K and Guerra V (1995) Attention focusing facilitatedthrough remote mental interaction Journal of the American Society for Psychical Research89 103

Cohen J (1997) Statistical Power Analysis for the Behavioral Sciences New York AcademicPress

Dalton K (1994) Remotely influenced ESP performance in a computer task A preliminarystudy Proceedings of the 37th Annual Convention of the Parapsychological Association pp95ndash103

Dossey L (1994) Healing and the mind Is there a dark side Journal of Scientific Exploration 873

Halifax-Grof J (1974) Hex death In A Angoff amp D Barth (Eds) Parapsychology and Anthro-pology Proceedings of an International Conference Held in London England August 29ndash311973 NY Parapsychology Foundation

Hope M (1983) Practical Techniques of Psychic Self-Defence Wellingborough Northampton-shire Aquarian Press

Marks DF (1973) Visual imagery differences in the recall of pictures British Journal of Psy-chology 64 17

Nash CB (1982) Psychokinetic control of bacterial growth Journal of the Society for PsychicalResearch 51 217

Radin DI Taylor RK and Braud WG (1995) Remote mental influence of human electroder-mal activity A pilot replication European Journal of Parapsychology 11 19

Schlitz M amp Braud W G (1997) Distant intentionality and healing Assessing the evidence Al-ternative Therapies 3 62

Schlitz M J amp LaBerge S (1997) Covert observation increases skin conductance in subjectsunaware of when they are being observed A replication Journal of Parapsychology 61 185

Schmeidler G R (1958) Agent-percipient relationships Journal of the American Society forPsychical Research 52 47

Schmeidler G R (1961) Evidence for two kinds of telepathy International Journal of Parapsy-chology 3 5

Tart C T (1979) A survey of expert opinion on potentially negative uses of psi United Statesgovernment interest in psi and the level of research funding of the field In WG Roll (Ed) Re-search in Parapsychology 1978 Metuchen NJ Scarecrow Press pp 54

Tart C T amp Labore C M (1986) Attitudes towards strongly functioning psi A preliminary sur-vey Journal of the America Society for Psychical Research 80 163

Tart C T (1986) Psychicsrsquo fears of psychic powers Journal of the American Society for Psychi-cal Research 80 278

Watt C A McDermott Z amp Ravenscroft J (1997) Exploring the limits of remote mental inter-action A comparison of blocking versus co-operating strategies In R Wiseman (Ed) Pro-ceedings of the 40th Annual Convention of the Parapsychological Association held in conjunc-tion with the Society for Psychical Research Brighton England pp 464ndash476

Blocking vs Co-operating 535

- 06) In Study 2 effect size measures indicated that the experimenters dif-fered in remote influence effect sizes (r = +17 and +30) Due to the smaller Nthis difference was not found to be statistically significant Post hoc combin-ing Studies 1 and 2 for experimenters JW and CW (ZMcD was not an experi-menter in Study 2) the effect sizes are r = +08 and r = +20 respectively

We conducted a post hoc analysis of the effects of influenceeinfluencer sexpairings on remote influence scores Our two studies came up with contradic-tory findings each statistically significant Study one found that mixed sexpairings had significantly higher remote influence scores than same sex pair-ings This replicated the post hoc finding of Schlitz and LaBerge (1997) in con-nection with the autonomic detection of remote observation Study 2 foundthat same sex pairings had significantly higher remote influence scores thanmixed sex pairings Our contradictory findings do not help to clarify the pic-ture on the importance or otherwise of sex pairings It is quite possible that sexis not the most salient aspect of the dynamic between pairs Age personalityand familiarity are likely to be important variables moderating the relationshipbetween influencer and influencee

There are many more variables that we could look at but as the number ofanalyses multiplies so does the opportunity for spurious significant resultsFor this reason we have named only two formal hypotheses for each experi-ment All other analyses are regarded as exploratory and any trends that wehave highlighted above would require replication

In conclusion these two studies have found no evidence of a difference inremote influence effects when influencees are asked to adopt cooperative men-tal strategies compared to when they are asked to adopt blocking mental strate-gies In fact contrary to expectation both studies found slightly higher remoteinfluence effects for the blocking condition compared to the co-operating con-dition On this basis we have found no indication of limiting conditions for re-mote influence effects Therefore while anecdotal advice exists on how to de-fend oneself against a remote physiological influence attempt in the realworld laboratory evidence to date has provided no support for this advice

Given the importance of this question both in practical terms and in theoret-ical terms we argue that this line of research deserves continued investigationIf a way could be found to conduct research with a more realistic scenario mdash where the remote influence was not just neutrally and abstractlydefined as unwelcome mdash then our research might have more meaningful thingsto say about the parameters of remote influence It is interesting and perhapssurprising to note that when Watt has lectured on this topic to informed lay au-diences the feedback has often been that researchers must use more potential-ly harmful or disruptive remote influence (of course with the informed consentof the participants) While there are clear ethical inhibitions when consideringpossible physiological influence perhaps research using higher level influencewould be more acceptable For example one could attempt to disrupt or en-hance the concentration of chess players who were aware of the influence

Blocking vs Co-operating 533

534 C Watt et al

attempt and who were concurrently blocking or co-operating with the influ-ence attempt Research already exists to suggest that remote influence does notjust operate at physiological levels Attention focusing for example can be fa-cilitated by a remote helper (Brady amp Morris 1997 Braud et al 1995) Futureresearch in this area must therefore expand beyond physiological influences

Acknowledgements

The authors gratefully acknowledge the financial support they receivedfrom the Institut fuumlr Grenzgebiete der Psychologie und PsychohygieneFreiburg Germany while conducting these studies We would also like tothank Dr Paul Stevens for generously helping to solve software difficultieswhen preparing for these studies and the anonymous reviewer for helpfulcomments Address reprint requests to Dr C Watt Department of PsychologyUniversity of Edinburgh 7 George Square Edinburgh EH8 9JZ Scotland UK

ReferencesBrady C amp Morris R L (1997) Attention focusing facilitated through remote mental interac-

tion A replication and exploration of parameters In R Wiseman (Ed) Proceedings of Pre-sented Papers The Parapsychological Association 40th Annual Convention held in conjunc-tion with The Society for Psychical Research pp 73ndash91

Braud W G (1985) Blockingshielding psychic functioning through psychological and psychictechniques A report of three preliminary studies In RA White and J Solfvin (Eds) Re-search in Parapsychology 1984 Metuchen NJ Scarecrow Press pp 42ndash44

Braud W G amp Schlitz M J (1991) Consciousness interactions with remote biological systemsAnomalous intentionality effects Subtle Energies 2 1

Braud W G Schlitz M J Collins J and Klitch H (1985) Further studies of the bio-PK effectFeedback blocking specificitygenerality In RA White amp J Solfvin (Eds) Research inParapsychology 1984 Metuchen NJ Scarecrow Press pp 45ndash48

Braud W G Shafer D McNeill K and Guerra V (1995) Attention focusing facilitatedthrough remote mental interaction Journal of the American Society for Psychical Research89 103

Cohen J (1997) Statistical Power Analysis for the Behavioral Sciences New York AcademicPress

Dalton K (1994) Remotely influenced ESP performance in a computer task A preliminarystudy Proceedings of the 37th Annual Convention of the Parapsychological Association pp95ndash103

Dossey L (1994) Healing and the mind Is there a dark side Journal of Scientific Exploration 873

Halifax-Grof J (1974) Hex death In A Angoff amp D Barth (Eds) Parapsychology and Anthro-pology Proceedings of an International Conference Held in London England August 29ndash311973 NY Parapsychology Foundation

Hope M (1983) Practical Techniques of Psychic Self-Defence Wellingborough Northampton-shire Aquarian Press

Marks DF (1973) Visual imagery differences in the recall of pictures British Journal of Psy-chology 64 17

Nash CB (1982) Psychokinetic control of bacterial growth Journal of the Society for PsychicalResearch 51 217

Radin DI Taylor RK and Braud WG (1995) Remote mental influence of human electroder-mal activity A pilot replication European Journal of Parapsychology 11 19

Schlitz M amp Braud W G (1997) Distant intentionality and healing Assessing the evidence Al-ternative Therapies 3 62

Schlitz M J amp LaBerge S (1997) Covert observation increases skin conductance in subjectsunaware of when they are being observed A replication Journal of Parapsychology 61 185

Schmeidler G R (1958) Agent-percipient relationships Journal of the American Society forPsychical Research 52 47

Schmeidler G R (1961) Evidence for two kinds of telepathy International Journal of Parapsy-chology 3 5

Tart C T (1979) A survey of expert opinion on potentially negative uses of psi United Statesgovernment interest in psi and the level of research funding of the field In WG Roll (Ed) Re-search in Parapsychology 1978 Metuchen NJ Scarecrow Press pp 54

Tart C T amp Labore C M (1986) Attitudes towards strongly functioning psi A preliminary sur-vey Journal of the America Society for Psychical Research 80 163

Tart C T (1986) Psychicsrsquo fears of psychic powers Journal of the American Society for Psychi-cal Research 80 278

Watt C A McDermott Z amp Ravenscroft J (1997) Exploring the limits of remote mental inter-action A comparison of blocking versus co-operating strategies In R Wiseman (Ed) Pro-ceedings of the 40th Annual Convention of the Parapsychological Association held in conjunc-tion with the Society for Psychical Research Brighton England pp 464ndash476

Blocking vs Co-operating 535

534 C Watt et al

attempt and who were concurrently blocking or co-operating with the influ-ence attempt Research already exists to suggest that remote influence does notjust operate at physiological levels Attention focusing for example can be fa-cilitated by a remote helper (Brady amp Morris 1997 Braud et al 1995) Futureresearch in this area must therefore expand beyond physiological influences

Acknowledgements

The authors gratefully acknowledge the financial support they receivedfrom the Institut fuumlr Grenzgebiete der Psychologie und PsychohygieneFreiburg Germany while conducting these studies We would also like tothank Dr Paul Stevens for generously helping to solve software difficultieswhen preparing for these studies and the anonymous reviewer for helpfulcomments Address reprint requests to Dr C Watt Department of PsychologyUniversity of Edinburgh 7 George Square Edinburgh EH8 9JZ Scotland UK

ReferencesBrady C amp Morris R L (1997) Attention focusing facilitated through remote mental interac-

tion A replication and exploration of parameters In R Wiseman (Ed) Proceedings of Pre-sented Papers The Parapsychological Association 40th Annual Convention held in conjunc-tion with The Society for Psychical Research pp 73ndash91

Braud W G (1985) Blockingshielding psychic functioning through psychological and psychictechniques A report of three preliminary studies In RA White and J Solfvin (Eds) Re-search in Parapsychology 1984 Metuchen NJ Scarecrow Press pp 42ndash44

Braud W G amp Schlitz M J (1991) Consciousness interactions with remote biological systemsAnomalous intentionality effects Subtle Energies 2 1

Braud W G Schlitz M J Collins J and Klitch H (1985) Further studies of the bio-PK effectFeedback blocking specificitygenerality In RA White amp J Solfvin (Eds) Research inParapsychology 1984 Metuchen NJ Scarecrow Press pp 45ndash48

Braud W G Shafer D McNeill K and Guerra V (1995) Attention focusing facilitatedthrough remote mental interaction Journal of the American Society for Psychical Research89 103

Cohen J (1997) Statistical Power Analysis for the Behavioral Sciences New York AcademicPress

Dalton K (1994) Remotely influenced ESP performance in a computer task A preliminarystudy Proceedings of the 37th Annual Convention of the Parapsychological Association pp95ndash103

Dossey L (1994) Healing and the mind Is there a dark side Journal of Scientific Exploration 873

Halifax-Grof J (1974) Hex death In A Angoff amp D Barth (Eds) Parapsychology and Anthro-pology Proceedings of an International Conference Held in London England August 29ndash311973 NY Parapsychology Foundation

Hope M (1983) Practical Techniques of Psychic Self-Defence Wellingborough Northampton-shire Aquarian Press

Marks DF (1973) Visual imagery differences in the recall of pictures British Journal of Psy-chology 64 17

Nash CB (1982) Psychokinetic control of bacterial growth Journal of the Society for PsychicalResearch 51 217

Radin DI Taylor RK and Braud WG (1995) Remote mental influence of human electroder-mal activity A pilot replication European Journal of Parapsychology 11 19

Schlitz M amp Braud W G (1997) Distant intentionality and healing Assessing the evidence Al-ternative Therapies 3 62

Schlitz M J amp LaBerge S (1997) Covert observation increases skin conductance in subjectsunaware of when they are being observed A replication Journal of Parapsychology 61 185

Schmeidler G R (1958) Agent-percipient relationships Journal of the American Society forPsychical Research 52 47

Schmeidler G R (1961) Evidence for two kinds of telepathy International Journal of Parapsy-chology 3 5

Tart C T (1979) A survey of expert opinion on potentially negative uses of psi United Statesgovernment interest in psi and the level of research funding of the field In WG Roll (Ed) Re-search in Parapsychology 1978 Metuchen NJ Scarecrow Press pp 54

Tart C T amp Labore C M (1986) Attitudes towards strongly functioning psi A preliminary sur-vey Journal of the America Society for Psychical Research 80 163

Tart C T (1986) Psychicsrsquo fears of psychic powers Journal of the American Society for Psychi-cal Research 80 278

Watt C A McDermott Z amp Ravenscroft J (1997) Exploring the limits of remote mental inter-action A comparison of blocking versus co-operating strategies In R Wiseman (Ed) Pro-ceedings of the 40th Annual Convention of the Parapsychological Association held in conjunc-tion with the Society for Psychical Research Brighton England pp 464ndash476

Blocking vs Co-operating 535

Schmeidler G R (1958) Agent-percipient relationships Journal of the American Society forPsychical Research 52 47

Schmeidler G R (1961) Evidence for two kinds of telepathy International Journal of Parapsy-chology 3 5

Tart C T (1979) A survey of expert opinion on potentially negative uses of psi United Statesgovernment interest in psi and the level of research funding of the field In WG Roll (Ed) Re-search in Parapsychology 1978 Metuchen NJ Scarecrow Press pp 54

Tart C T amp Labore C M (1986) Attitudes towards strongly functioning psi A preliminary sur-vey Journal of the America Society for Psychical Research 80 163

Tart C T (1986) Psychicsrsquo fears of psychic powers Journal of the American Society for Psychi-cal Research 80 278

Watt C A McDermott Z amp Ravenscroft J (1997) Exploring the limits of remote mental inter-action A comparison of blocking versus co-operating strategies In R Wiseman (Ed) Pro-ceedings of the 40th Annual Convention of the Parapsychological Association held in conjunc-tion with the Society for Psychical Research Brighton England pp 464ndash476

Blocking vs Co-operating 535