Severe violations of independence in response inhibition tasks

See allHide authors and affiliations

Science Advances  17 Mar 2021:
Vol. 7, no. 12, eabf4355
DOI: 10.1126/sciadv.abf4355
  • Fig. 1 Violations at short SSDs in experiments designed to evaluate violations.

    Violations across SSDs in the fixed SSD 1 (A) and fixed SSD 2 (B) conditions using linear mixed effects modeling (corrected for multiple comparisons across SSD values). Positive values indicate violations of context independence. Shaded areas indicate the 95% confidence interval. Fixed SSD 1 (A) includes SSDs between 100 and 500 ms at 100-ms increments, and fixed SSD 2 (B) includes SSDs between 0 and 500 ms at 50-ms increments.

  • Fig. 2 Severe violations of independence.

    The thick black line is the mean across studies, and colored lines are individual conditions or studies. (A) Positive values indicate severe violations. The gray band around the mean is the 95% confidence interval of the results of the linear mixed effect model using data from all conditions and studies (corrected for multiple comparisons across SSD values). Figure S2 is identical but includes a legend. (B) Proportion of conditions that violate [i.e., positive values from (A) at each SSD]. (C) Proportion of individual subjects from each condition that violate the independence assumption at each SSD. Figure S3 is identical but includes a legend. (D) Cumulative proportion of SSDs in our 23 conditions that tracked SSD with a 1 up 1 down tracking algorithm. Note: A small proportion of stop trials had negative SSDs (0.02) or SSDs above 750 ms (0.007) and are not displayed here or included in any analyses of violations. Figure S7 is identical but includes a legend.

  • Fig. 3 Fundamental conclusions can change when short SSDs are removed.

    Note: Error bars in (B) and (C) represent 95% confidence intervals from bootstrapping. (A) SSRT is significantly faster in 24 of our 25 conditions when short SSDs (< 200 ms) were removed. Figure S8 is identical but includes a legend. (B) SSRT is only faster in the 40% stop signal than the 20% stop signal condition in subjects with predominantly shorter (mean < 300 ms) SSDs. (C) Violations are present at shorter SSDs (<250 ms; left) but not at longer SSDs (>250 ms; right) for all subjects in stimulus selective stopping, bringing into question the putative differences in strategies (SD, stop then discriminate; DDS, dependent discriminate then stop) proposed by Bissett and Logan (27).

  • Table 1 Basic information on analyzed datasets and conditions.

    N is the number of subjects before excluding based on an insufficient number of trials at short SSDs (see the Supplementary Materials for details). Trial N is the number of total trials per subject.

    Condition no.Condition
    NTrial NCollection
    Stop modalityStop rule(Citation no.)
    1Fixed SSD 1241200In personAuditorySimple(36)
    2Fixed SSD 2242400In personAuditorySimple(36)
    3Deadline 3
    (300 ms)
    24480In personAuditorySimple(36)
    4Deadline 1
    (500 ms)
    480In personAuditorySimple(36)
    5Deadline 1
    (700 ms)
    480In personAuditorySimple(36)
    6Deadline 2
    (300 ms)
    24480In personAuditorySimple(36)
    7Deadline 2
    (500 ms)
    480In personAuditorySimple(36)
    8Deadline 2
    (700 ms)
    480In personAuditorySimple(36)
    9Stop probability 0.2241200In personAuditorySimple(21), E. 1
    10Stop probability 0.41200In personAuditorySimple(21), E. 1
    11Turk simple 0.2339300Mechanical TurkVisualSimple(40)
    12Turk simple 0.4300Mechanical TurkVisualSimple(40)
    13Turk stim
    300Mechanical TurkVisualStim selective(40)
    14Turk motor
    300Mechanical TurkVisualMotor selective(40)
    15Saccadic eye
    11600In personAuditorySimple(36)
    modality auditory 1
    241200In personAuditorySimple(42) E. 1
    modality auditory 2
    241200In personAuditorySimple(42) E. 2
    modality visual 1
    321200In personVisualSimple(42) E. 3
    modality visual 2
    241200In personVisualSimple(42) E. 4
    stimulus selective
    241200In personAuditoryStim selective(27) E. 1
    central go simple
    24520In personAuditorySimple(36)
    central go
    selective stop
    520In personAuditoryStim selective(36)
    peripheral go
    simple stop
    520In personAuditorySimple(36)
    peripheral go
    selective stop
    520In personAuditoryStim selective(36)
    25Variable difficulty53576In personVisualSimple(38)
  • Table 2 Which variables influence the violation?

    Most variables do not affect the violation (although note that Bayes factors (BF10) in support of the null hypothesis in rows 1 and 3 to 5 were approximately 3, suggesting anecdotal evidence), but extremely short deadlines (row 2) reduce the violation and introducing stimulus or motor selectivity increases the violation (rows 6 to 8, though note equivocal Bayes factors). Numbers in parentheses in the condition column correspond to the condition column in Table 1. Statistics were based on the analysis of variance (ANOVA) interaction of the trial type (preceding no stop versus stop fail) and the condition on mean RT. ƞ2 is a measure of effect size

    1Fast vs. slow
    RT below median
    (1/2) vs. RT above
    media (1/2)
    6 vs. 15 ms0.5493.0340.0009846
    2Varying go
    response deadlines
    300 ms (3/6) vs. 500 ms
    (4/7) vs. 700 ms
    −16 vs. −5 vs. 3 ms0.0112.4780.008
    3Low vs. high stop
    0.2 (9/11) vs. 0.4 stop
    probability (10/12)
    8 vs. 27 ms0.1272.5820.007
    4Saccadic vs. manual
    Saccadic (15) vs.
    manual responses (9)
    −9 vs. −3 ms0.7543.7999.382 × 10−5
    5Auditory vs. visual
    Auditory (16/17) vs.
    visual responses
    −22 vs. −13 ms0.6373.0950.0007657
    6Stimulus selective
    vs. simple stopping
    Stimulus selective (20)
    vs. simple stopping (9)
    55 vs. −1 ms0.0640.8020.018
    7Stimulus selective
    vs. simple stopping
    Stimulus selective
    (13/22/24) vs.
    simple stopping (9)
    49 vs. 24 ms0.0180.7940.014
    8Motor selective vs.
    simple stopping
    Motor selective (14)
    vs. simple stopping
    69 vs. 36 ms0.0191.4670.015

Supplementary Materials

  • Supplementary Materials

    Severe violations of independence in response inhibition tasks

    Patrick G. Bissett, Henry M. Jones, Russell A. Poldrack, Gordon D. Logan

    Download Supplement

    This PDF file includes:

    • Supplementary Text
    • Figs. S1 to S9
    • References

    Files in this Data Supplement:

Stay Connected to Science Advances

Navigate This Article