agape experiment: further statistical studies (in progress) dr bernard auriol (europa meeting,...
TRANSCRIPT
Agape Experiment:Further Statistical
Studies(in progress)
Dr Bernard Auriol
(EuroPA meeting, November 2003)
ProtocolAgape Experiment lasted seven years to test H1 :
The rate of hits could be increased by redundancy due to vote. To test it, there were a transmitting group (emission-strength of 1 to 16 senders) and a receiving group (1 to 16 voters), located in two separate isolate rooms. Everything was monitored and recorded thanks to several computers and a network especially designed for the experiment.
Three main protocols were tested with Two pictures , Three words, or Five words as possible target. Moreover, different parameters varied from one session to another in order to find the best conditions for later replications.
240 telepathic ESP group sessions,
27,845 collective trials
250,000 individual trials !
Participants : any voluntary
either sheep or goat
sender or receiver role generally chosen by the participants
274 female (2/3)
145 male (1/3)
target’s type : either pictures or words
possible answers’ number => 2, 3 or 5
Individual Answers
Nb of tries Success P-value
2 pictures 27,081 49.94% 0.83
3 words 102,634 33.34% 0.95
5 words 120,347 20.13% 0.27
Majority VoteMajority Vote
Possible Targets
Expected Mean
Observed Mean
2 images 0.500 0.498
3 words 0.333 0.329
5 words 0.200 0.202
Targets Nb Salvos Obs.Var. Exp.Var.
2 196 3.08* 3.75
3 713 3.37 3.33
5 268 2.39 2.40
Variance of success got Variance of success got by voteby vote
(15 trials per salvo)
Targets Nb Salvos Obs.Var. Exp.Var.
2 98 5.26* 7.50
3 356 6.32 6.67
5 134 4.72 4.80
Variance of success got Variance of success got by voteby vote
(30 trials per salvo)
Targets Nb Salvos Obs.Var. Exp.Var.
2 49 11.92* 15.00
3 178 12.39 13.33
5 67 _9.03 _9.60
Variance of success got Variance of success got by voteby vote
(60 trials per salvo)
Variance of intervalsVariance of intervals
To reach a better evaluation, we note the interval ( number of misses)
between two consecutive hits, and check the variance of these intervals
(at random or not ?).
Variance of the Variance of the intervalsintervals
Targets Success Obs. Var. Exp. Var.
2 1463 1.93* _2.00
3 3512 6.42* _6.00
5 _777 20.05 20.00
Instable Attitudes of Receivers ?
That apparently inconsistent set of variances
shifting according to the protocol
could be linked to an alternation along the sessions
of Goal-Oriented Socio-psychological Attitudes
producing in turn
Psi-Nothing, Psi-Hitting and Psi-Missing.
Majority Strength and Majority Strength and
Success of VoteSuccess of VoteIf some answers are not due to chance but to ESP,
this should have an impact on the majority: Strong majorities could be more linked to success than weak ones.
Unlike what we expected, strong majorities didn’t get better results than weak ones
ConclusionConclusionof the hypothesis test
The results did not fulfil our hope regarding a possible improvement of the Signal to Noise ratio (S/N)
(redundancy got from majority vote).
This way of carrying out the experiment, did not strongly increase the Psi-Hitting rate as we expected,
but seems to have made the results random, regarding either individual answers or answers obtained by vote.
Prospective purpose:
Covariance Analysis
Nevertheless, for heuristic purpose,we undertook a covariance analysis on collective trials which are significantly different from chance (p < 0.05)
In order to test the effect of each variables modality, we used a transformation of the « percentage of hits »
to be able to compare the results for the protocols with two pictures, three or five words.
where:- is the percentage of right answers in the trial
p is the expected percentage
We can test if the percentage of hits got for each tria l issignificantly lower than chance, significantly higher than chance,
or equal to chance,thanks to a test of Khi2.
The statistic of this test, calculated for each trial, is
with po = expected percent of successes; = observed percent of successes
~Khi2 (one df)
Collective trials Collective trials significantly significantly differentdifferent from chance (p <0.05)from chance (p <0.05)
Nb. of tries Percent of tries
Higher than chance __1079 _3.87
Lower than chance ___413 _1.48
Equal to chance 26.353 94.64
We focused on the trials where the percentage of hits differed significantly from chance expectation.
The effect of different parameters on the answers kept was outlined with a covariance analysis.
The significant variables were selected thanks to a stepwise procedure
and kept under a threshold of 5%.
We get significant parameters with a p-value close to 0.0001
Collective trials significantly departing from chance
Estimate Higher than chance Lower than
chance Constant + 1.06 - 1.12
Instruction given to transmitters
+ 0.10 …
Targets’ list chosen - 0.34 + 0.26 Relevant reward - 0.13 …
The time left once half the receivers answered equals the time already
spent
+ 0.27 - 0.02
(…) twice the time already spent
+ 0.30 - 0.12
Qualitative Variables
Collective trials significantly departing from chance
Estimate Higher than chance Lower than chance
Constant + 1.06 - 1.12 Time to answer - 0.01 + 0.01
Ratio nb. of transmitters / nb. of participants
- 1.46 …
Ratio nb. of transmitters / nb. of receivers
+ 0.96 …
Quantitative Variables
Effect of the relationship Effect of the relationship between participantsbetween participants
Variable Mark given to
Estimate Success Distance from chance
P-value
RecGr Receivers -1,54 < 0.0001 TransGr Senders +0,58 __0.0003
Familiarity between Receivers moves the results closer to chance.
Familiarity between Senders and Receivers moves the results away from chance.
For the trials significantly higher than chance expectation,adjusted R² equals 10.8% (« small » effect according to Cohen’s convention)
Conclusion
If there was any Psi manifestation in our experiment,there would be a strong tendency to reject the right answeras an aggression coming from the transmitters’ sub-group.
Couldn’t that be the need for each individualto avoid his own dissolution,
especially if the individual belongs to a group,(situation which favours the fusion among the members) ?
Is Psi-Missing a cross-boarder defense system
?
Sybil
We plan to devise a protocol to test the following hypothesis:
We can hope for success with groups only if we build sub-groups so that there is more affinity between receivers and senders than among receivers. A simple sociometric test should be enough to achieve this, provided the results for each sessions help to distribute the roles of transmitter and receiver.