426_2016_795_moesm1_esm.docx - springer static …10.1007/s004…  · web viewm = 120 ms; t [24] =...

17
Experiment 2B Experiment 2B served two goals. The setup was modeled after Experiment 1 to first, provide an independent replication of the findings of PM-cue-category similarity (Experiment 1) and second, to test whether the observed erroneous activation of the current PM response by PM REPEATED trials in Experiment 1 and 2 depended on the manual modality of the currently relevant PM response. Experiment 2B was therefore similar to Experiment 1, the only difference being that participants now responded verbally in the PM-response-switch condition. Consequently, not only the content of the PM response but also its modality (i.e., manual versus verbal) differed between the PM block and the test block. Method Participants. Twenty-five students (10 male; age M = 23.40 years, SD = 3.29) who had not participated in Experiment 1 or Experiment 2 took part in Experiment 2B. Apparatus and stimuli. Manual responses on the Y key (Experiment 1) were replaced by verbal responses (i.e., pronouncing “Ja” [German = Yes]), which were recorded via voice key using Presentation software (Version 16.1; www.neurobs.com).

Upload: vudien

Post on 06-Feb-2018

213 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: 426_2016_795_MOESM1_ESM.docx - Springer Static …10.1007/s004…  · Web viewM = 120 ms; t [24] = 6.41 ... Errors might also occur when performing the ongoing task on oddball and

Experiment 2B

Experiment 2B served two goals. The setup was modeled after Experiment 1 to first,

provide an independent replication of the findings of PM-cue-category similarity (Experiment

1) and second, to test whether the observed erroneous activation of the current PM response

by PMREPEATED trials in Experiment 1 and 2 depended on the manual modality of the currently

relevant PM response. Experiment 2B was therefore similar to Experiment 1, the only

difference being that participants now responded verbally in the PM-response-switch

condition. Consequently, not only the content of the PM response but also its modality (i.e.,

manual versus verbal) differed between the PM block and the test block.

Method

Participants. Twenty-five students (10 male; age M = 23.40 years, SD = 3.29) who

had not participated in Experiment 1 or Experiment 2 took part in Experiment 2B.

Apparatus and stimuli. Manual responses on the Y key (Experiment 1) were replaced

by verbal responses (i.e., pronouncing “Ja” [German = Yes]), which were recorded via voice

key using Presentation software (Version 16.1; www.neurobs.com).

Procedure and design. The experimental procedure was analogous to Experiment 1

with the exception that participants were required to verbally respond (instead of pressing the

Y key) to PM cues in the test block in the PM-response-switch condition.

Results

Error trials (5.2%) and trials with RTs 2.5 SDs above or below a participant’s mean

RT for a given trial type (PM block: 2.5%; test block: 2.9%) were excluded prior to RT

analyses. Results are presented in Figure 1 and Table 1.

PM block. We found an orientation response on oddball trials (M = 748 ms, SD = 92

ms; M = 8.8%, SD = 5.1%) in terms of increased RTs and error rates compared to standard

trials (M = 549 ms, SD = 54 ms; M = 3.7%, SD = 2.3%), t(24) = 19.24, p < .001, d = 2.65; and

Page 2: 426_2016_795_MOESM1_ESM.docx - Springer Static …10.1007/s004…  · Web viewM = 120 ms; t [24] = 6.41 ... Errors might also occur when performing the ongoing task on oddball and

t(24) = 7.02, p < .001, d = 1.29, respectively. RTs (M = 639 ms, SD = 74 ms) and error rates

(M = 8.7%, SD = 5.6%) on PMSYMBOL trials were comparable to Experiments 1.

Figure 1. Results Experiment 2B. Mean response time (RT) and percent error as a function of trial type (prospective memory [PM]SYMBOL, standard, oddball) in the PM block and as a function of PM-response similarity (repetition [PM block: X key; test block: X key], switch [PM block: X key, test block: voice key]) and PM-cue-category similarity (repetition [PM block: symbol; test block: symbol], switch [PM block: symbol; test block: word]) and trial type (PMSYMBOL/PMWORD, standard, oddball, PMREPEATED) in the test block. B) Mean percentage erroneous X-key and voice-key responses as a function of PM-response similarity and PM-cue-category similarity and trial type (oddball, PMREPEATED) in the test block. Error bars represent standard errors.

Test block.

Aftereffects of completed intentions. We analyzed aftereffects by conducting 2 (trial

type: PMREPEATED trials, oddball trials) × 2 (PM-cue-category similarity: repetition, switch) × 2

(PM-response similarity: repetition, switch) repeated measures ANOVAs on RTs of the

ongoing task and on commission errors (i.e., erroneous X-key presses and voice key

responses).

RTs. We found aftereffects in terms of increased RTs on PMREPEATED trials (M = 878

ms) compared to oddball trials (M = 776 ms), F(1, 24) = 58.92, p < .001, ηp2 = .71. RTs were

slower in the PM-cue-category-repetition condition (M = 889 ms) than in the PM-cue-

category-switch condition (M = 765 ms), F(1, 24) = 122.25, p < .001, ηp2 = .84. Aftereffects

Page 3: 426_2016_795_MOESM1_ESM.docx - Springer Static …10.1007/s004…  · Web viewM = 120 ms; t [24] = 6.41 ... Errors might also occur when performing the ongoing task on oddball and

were larger in the PM-cue-category-repetition condition (M = 120 ms; t[24] = 6.41, p < .001,

d = 1.00) than in the PM-cue-category-switch condition (M = 84 ms; t[24] = 7.17, p < .001, d

= 0.70), as shown by a Trial type × PM-cue-category similarity interaction, F(1, 24) = 4.83, p

= .038, ηp2 = .17. RTs were slightly faster in the PM-response-repetition condition (M = 819

ms) than in the PM-response-switch condition (M = 836 ms), F(1, 24) = 4.31, p = .049, ηp2

= .15 (other Fs < 1.32, ps > .263).

Commission errors. Indicating aftereffects, the rate of erroneous X key responses was

increased on PMREPEATED trials (M = 2.8%) compared to oddball trials (M = 0.4%), F(1, 24) =

31.22, p < .001, ηp2 = .56. Participants made more errors in the PM-cue-category-repetition

condition (M = 2.4%) than in the PM-cue-category-switch condition (M = 0.8%), F(1, 24) =

12.81, p = .002, ηp2 = .35. Errors were increased in the PM-response-repetition condition (M =

3.0%) compared to the PM-response-switch condition (M = 0.3%), F(1, 24) = 34.92, p < .001,

ηp2 = .59. X key aftereffects differed between conditions, as revealed by Trial type × PM-cue-

category-similarity, F(1, 24) = 6.83, p = .015, ηp2 = .22; Trial type × PM-response-similarity,

F(1, 24) = 21.75, p < .001, ηp2 = .48; and Trial type × PM-cue-category-similarity × PM-

response-similarity interactions, F(1, 24) = 4.78, p = .039, ηp2 = .17. That is, aftereffects were

only significant in the PM-cue-category-repetition/PM-response-repetition condition (M =

6.2%; t[24] = 5.02, p < .001, d = 1.29) and in the PM-cue-category-switch/PM-response-

repetition condition (M = 2.5%; t[24] = 2.68, p = .013, d = 0.79), but not in the PM-cue-

category-repetition/PM-response-switch condition (M = 0.8%; t[24] = 2.00, p = .057, d =

0.56) and in the PM-cue-category-switch/PM-response-switch condition (M = 0.2%; t[24] =

1.00, p = .327, d = 0.28).

The similar analysis on erroneous voice key responses revealed a slightly increased

error rate in the PM-cue-category-repetition condition (M = 0.8%) compared to the PM-cue-

category-switch condition (M = 0.3%), F(1, 24) = 7.58, p = .011, ηp2 = .24. Participants made

more errors in the PM-response-switch condition (M = 0.8%) compared to the PM-response-

Page 4: 426_2016_795_MOESM1_ESM.docx - Springer Static …10.1007/s004…  · Web viewM = 120 ms; t [24] = 6.41 ... Errors might also occur when performing the ongoing task on oddball and

repetition condition (M = 0.2%), F(1, 24) = 6.68, p = .016, ηp2 = .22. Most important and in

contrast to Y-key errors in Experiment 1, trial type and all further interactions were not

significant, Fs < 1.87, ps > .185.

Table 1Mean RTs and Ongoing-Task Errors for the Test block and ongoing-task errors by PM-Cue-Category Similarity, PM-Response Similarity and Trial Type in Experiment 2B (SDs in Parentheses).

RT (ms) Ongoing-task error (%)1

PM-cue-category repetition (symbol)

Response repetition(X key)

PMSYMBOL 679 (75) 9.5 (9.0)Standard 625 (62) 4.1 (2.6)Oddball 808 (78) 5.8 (6.0)PMREPEATED 948 (148) 9.0 (5.9)

Response switch (voice key)

PMSYMBOL 805 (125) 15.0 (13.7)Standard 629 (62) 4.8 (3.2)Oddball 851 (117) 8.0 (5.8)PMREPEATED 951 (162) 11.0 (8.1)

PM-cue-categoryswitch (word)

Response repetition(X key)

PMWORD 724 (116) 11.2 (8.3)Standard 635 (62) 4.0 (2.6)Oddball 719 (121) 6.0 (5.9)PMREPEATED 801 (124) 4.7 (5.0)

Response switch (voice key)

PMWORD 818 (119) 14.7 (10.6)Standard 637 (60) 4.8 (3.0)Oddball 728 (94) 7.7 (7.3)PMREPEATED 813 (153) 6.3 (7.7)

Note. 1 Error rates on PM trials are omission errors.

Ongoing-task errors. For ongoing-task errors (i.e., performing an incorrect word

response), we did not find overall aftereffects, F(1, 24) = 0.90, p = .353, ηp2 = .03. More errors

occurred in the PM-cue-category-repetition condition (M = 8.5%) than in the PM-cue-

category-switch condition (M = 6.2%), F(1, 24) = 8.75, p = .007, ηp2 = .27. Trial type and PM-

cue-category similarity interacted, F(1, 24) = 11.59, p = .002, ηp2 = .33. Specifically, we found

aftereffects in the PM-cue-category-repetition condition (M = 3.1%; t[24] = 3.00, p = .006, d

= 0.55) but not in the PM-cue-category-switch condition (M = -1.3%; t[24] = -1.09, p = .285,

d = -0.23). Finally, error rates were increased in the PM-response-switch condition (M =

8.3%) than in the PM-response-repetition condition (M = 6.4%), F(1, 24) = 7.25, p = .013, ηp2

= .23 (other Fs < 1, ps > .780).

Page 5: 426_2016_795_MOESM1_ESM.docx - Springer Static …10.1007/s004…  · Web viewM = 120 ms; t [24] = 6.41 ... Errors might also occur when performing the ongoing task on oddball and

Standard trials. Arguing for increased task difficulty and similar to Experiment 1, a 2

(PM-cue-category similarity: repetition, switch) × 2 (PM-response similarity: repetition,

switch) repeated-measures ANOVA revealed slower standard trial RTs in the PM-cue-

category-switch condition (M = 636 ms) than in the PM-cue-category-repetition condition (M

= 627 ms), F(1, 24) = 7.23, p = .012, ηp2 = .23 (other Fs < 2.32, ps > .142). The analogous

ANOVA for error rates showed that participants made slightly more errors in the PM-

response-switch condition (M = 5.0%) than in the PM-response-repetition condition (M =

4.3%), F(1, 24) = 5.56, p = .027, ηp2 = .19 (further Fs < 1, ps > .704).

PM trials. We computed 2 (PM-cue-category similarity: repetition [PM cues =

symbols], switch [PM cues = words]) × 2 (PM-response similarity: repetition [X key], switch

[voice key]) repeated-measures ANOVAs on RTs and errors on PM trials of the test block.

RTs in the PM-cue-category-repetition condition (PMSYMBOL: M = 742 ms) were again shorter

than in the PM-cue-category-switch condition (PMWORD: M = 771 ms), F(1, 24) = 4.34, p

= .048, ηp2 = .15. Verbal PM responses (M = 811 ms) in the PM-response-switch condition

were slower than manual responses (M = 702 ms) in the PM-response-repetition condition,

F(1, 24) = 36.46, p < .001, ηp2 = .60 (further F = 2.12, p = .158). Participants made fewer

errors when responding to PMSYMBOL cues (M = 10.3%, PM-cue-category-repetition condition)

than to PMWORD cues (M = 14.8%, PM-cue-category-switch condition), F(1, 24) = 6.31, p

= .019, ηp2 = .21 (further Fs < 1, ps > .586).

Discussion

With a new sample of participants we replicated the important finding from

Experiment 1 that aftereffects were affected by PM-cue-category similarity. That is, all types

of aftereffects (RTs, commission errors and even ongoing-task errors) were more pronounced

in the PM-cue-category-repetition condition compared to the PM-cue-category-switch

condition.

Page 6: 426_2016_795_MOESM1_ESM.docx - Springer Static …10.1007/s004…  · Web viewM = 120 ms; t [24] = 6.41 ... Errors might also occur when performing the ongoing task on oddball and

Most importantly, in Experiment 2B the PM-response-switch condition included a

switch from manual to verbal PM responses. Whereas RT aftereffects were not affected by

PM-response similarity (analogously to Experiment 1 and 2), error aftereffects showed a

differential pattern. PMREPEATED trials erroneously triggered current PM response execution

when PM responses remained the same in the PM block and the test block (i.e., both manual).

In PM-response-switch conditions, the formerly manual PM response did not trigger the

erroneous execution of the now verbal PM response upon encounter of PMREPEATED trials. This

finding might reflect an increased control over verbal response activation than over manual

response activation. Given faster manual than verbal response initiation, substantial parts of

the manual ongoing task response might already be underway before the erroneous verbal

response can be executed.

As a consequence, the finding that PMREPEATED trials erroneously trigger the activation

of the currently relevant PM response is not a generalized finding but seems to depend on the

modality of the current PM response (i.e., manual responses in the present experiment).

Page 7: 426_2016_795_MOESM1_ESM.docx - Springer Static …10.1007/s004…  · Web viewM = 120 ms; t [24] = 6.41 ... Errors might also occur when performing the ongoing task on oddball and

Analyses of Ongoing-Task Errors in Experiments 1 – 4

Errors might also occur when performing the ongoing task on oddball and PMREPEATED

trials but executing an incorrect word response. These errors should inform about ongoing-

task difficulty and/or cognitive resources available for the ongoing task (Table 2).

Experiment 1

In Experiment 1, in the corresponding 2 (trial type: PMREPEATED, oddball) × 2 (PM-cue-

category similarity: repetition, switch) × 2 (PM-response similarity: repetition, switch)

ANOVA, no evidence for an overall aftereffect was found for ongoing-task errors, F(1, 25) =

0.45, p = .510, ηp2 = .02. Ongoing-task errors were affected by PM-cue-category similarity,

F(1, 25) = 13.88, p = .001, ηp2 = .36, as participants made more errors in the PM-cue-category-

repetition condition (M = 9.5%) than in the PM-cue-category-switch condition (M = 6.6%).

PM-response-similarity was not significant, F(1, 25) = 0.02, p = .889, ηp2 = .00. However,

although trial type interacted with PM-response similarity, indicating that aftereffects were

especially pronounced in the PM-response-switch condition (M = 2.6%) as compared to the

PM-response-repetition condition (M = -1.3%), F(1, 25) = 6.73, p = .016, ηp2 = .21, t-tests on

differences between PMREPEATED trials and oddball trials did not reach significance (all ps

>.060). All further effects were not significant, Fs < 1, p > .608.

Experiment 2

In Experiment 2, the corresponding 2 (trial type: PMREPEATED, oddball) × 2 (PM-

response similarity: repetition, switch) repeated-measures ANOVA yielded a main effect of

trial type, F(1, 23) = 12.03, p = .002, ηp2 = .34, indicating increased ongoing task error rates on

PMREPEATED trials (M = 14.6%) compared to oddball trials (M = 10.9%). We neither observed a

main effect of PM-response similarity, F(1, 23) = 0.01, p = .942, ηp2 = .00, nor an interaction

between trial type and PM-response similarity, F(1, 23) = 2.23, p = .149, ηp2 = .09.

Page 8: 426_2016_795_MOESM1_ESM.docx - Springer Static …10.1007/s004…  · Web viewM = 120 ms; t [24] = 6.41 ... Errors might also occur when performing the ongoing task on oddball and

Experiment 3

In Experiment 3, the corresponding 2 (trial type: PMREPEATED, oddball) × 3 (condition:

PM-task repetition, PM-task switch, ongoing-task only) repeated-measures ANOVA revealed

that ongoing-task errors were affected by condition, F(2, 30) = 5.50, p = .012, ηp2 = .27.

Specifically, error rates were larger in the PM-task-repetition condition (10.6%) than in the

PM-task-switch condition (7.6%), F(1, 15) = 6.99, p = .018, ηp2 = .32; which in turn did not

differ from the ongoing-task-only condition (6.9%), F(1, 15) = 0.44, p = .516, ηp2 = .03

(repeated contrasts) (other Fs < 1, ps > .623).

Experiment 4

Mean ongoing-task errors were comparable to previous experiments (M = 10.5%).

However, the 2 (trial type: PMREPEATED, oddball) × 2 (PM-cue-category similarity: repetition,

switch) × 2 (PM-block-cue format: color, shape) ANOVA did not reveal significant effects,

Fs < 3.24, ps > .082.

Page 9: 426_2016_795_MOESM1_ESM.docx - Springer Static …10.1007/s004…  · Web viewM = 120 ms; t [24] = 6.41 ... Errors might also occur when performing the ongoing task on oddball and

Tables with Response Times and Ongoing-Task Errors in Experiments 1 – 4

Table 2Mean RTs and Ongoing-Task Errors for the Test block and ongoing-task errors by PM-Cue-Category Similarity, PM-Response Similarity and Trial Type in Experiment 1, 2 and 3 (SDs in Parentheses).

RT (ms) Ongoing-task error (%)1

Experiment 1

PM-cue-category repetition (symbol)

Response repetition(X key)

PMSYMBOL 705 (101) 7.7 (7.5)Standard 618 (94) 4.6 (3.4)Oddball 813 (136) 10.4 (7.8)PMREPEATED 966 (162) 8.5 (6.7)

Response switch(Y key)

PMSYMBOL 685 (90) 9.8 (6.9)Standard 625 (98) 4.7 (3.1)Oddball 711 (124) 6.9 (5.6)PMREPEATED 814 (170) 6.3 (6.8)

PM-cue-category switch(word)

Response repetition(X key)

PMWORD 694 (99) 14.3 (11.1)Standard 620 (97) 4.8 (3.1)Oddball 803 (152) 8.3 (5.4)PMREPEATED 949 (163) 10.9 (8.4)

Response switch(Y key)

PMWORD 689 (76) 16.2 (9.7)Standard 627 (104) 4.6 (2.8)Oddball 712 (108) 5.3 (6.9)PMREPEATED 796 (188) 7.9 (7.1)

Experiment 2Response repetition(voice key)

PMSYMBOL 885 (128) 16.1 (13.9)Standard 614 (72) 5.6 (3.7)Oddball 857 (160) 10.0 (7.1)PMREPEATED 1015 (179) 15.6 (7.9)

Response switch(X key)

PMSYMBOL 694 (91) 18.8 (23.0)Standard 618 (72) 6.0 (3.5)Oddball 837 (151) 11.7 (9.0)PMREPEATED 996 (169) 13.6 (8.8)

Experiment 3PM-task repetition(X key, symbol)

PMSYMBOL 674 (79) 14.6 (12.7)Standard 596 (87) 5.6 (3.9)Oddball 754 (165) 10.5 (8.8)PMREPEATED 909 (183) 10.7 (7.0)

PM-task switch(Y key, word)

PMWORD 678 (114) 17.8 (10.6)Standard 602 (91) 5.0 (2.9)Oddball 668 (117) 8.0 (5.6)PMREPEATED 759 (135) 7.2 (5.6)

Ongoing-task onlyStandard 583 (89) 5.6 (3.4)Oddball 630 (108) 7.2 (5.6)PMREPEATED 635 (108) 6.6 (7.0)

Note. 1 Error rates on PM trials are omission errors.

Page 10: 426_2016_795_MOESM1_ESM.docx - Springer Static …10.1007/s004…  · Web viewM = 120 ms; t [24] = 6.41 ... Errors might also occur when performing the ongoing task on oddball and

Table 3Mean RTs and Ongoing-Task Errors for the Test block and ongoing-task errors by PM-Cue-Category Similarity, PM-Response Similarity and Trial Type in Experiment 4 (SDs in Parentheses).

RT (ms) Ongoing-task error (%)1

PM-block format: color

PM-cue-category-repetition

PMCOLOR 705 (139) 18.1 (16.9)Standard 608 (96) 6.3 (4.9)Oddball 945 (207) 11.3 (11.1)PMREPEATED 959 (205) 9.4 (8.1)

PM-cue-category-switch

PMSHAPE 678 (109) 14.4 (13.2)Standard 600 (89) 6.2 (4.7)Oddball 877 (166) 12.8 (11.1)PMREPEATED 970 (193) 11.8 (11.2)

PM-block format: shape

PM-cue-category-repetition

PMSHAPE 664 (88) 12.2 (12.2)Standard 601 (92) 6.1 (4.6)Oddball 906 (165) 10.8 (9.0)PMREPEATED 986 (214) 9.0 (7.6)

PM-cue-category-switch

PMCOLOR 682 (107) 14.4 (13.3)Standard 603 (96) 6.4 (4.5)Oddball 896 (175) 9.9 (9.7)PMREPEATED 983 (200) 9.0 (9.6)

Note. 1 Error rates on PM trials are omission errors.

Page 11: 426_2016_795_MOESM1_ESM.docx - Springer Static …10.1007/s004…  · Web viewM = 120 ms; t [24] = 6.41 ... Errors might also occur when performing the ongoing task on oddball and

Stimulus Material Experiments 1, 2, 2B and 3Stimulus features serving as PMSYMBOL trials, PMREPEATED trials, oddball

trials in Experiments 1, 2, 2B and 3 are shown. Please note that each stimulus feature could appear with the ongoing-task digits 2 to 9 during the PM block and with food or non-food words during the Test block.

Page 12: 426_2016_795_MOESM1_ESM.docx - Springer Static …10.1007/s004…  · Web viewM = 120 ms; t [24] = 6.41 ... Errors might also occur when performing the ongoing task on oddball and