Ego Depletion: Is the Active Self a Limited Resource?
Journal of Personality and Social Psychology, 1998 · paper · Psychology
Four experiments test whether acts of self-control and choice draw on a shared limited resource. Initial self-regulation, such as resisting chocolate, reduced later persistence and performance and increased passivity, suggesting a common limited capacity for volition.
Checked: PDF from faculty.washington.edu, fetched 5 Oct 2026. Page numbers are those of that PDF.
0 findings
None that passed a second check.
9 minor (typos and the like; most checked by the judge only)
Wording For the desire-to-stop item, "F(2, 64) = 4.71, p < .01": the F statistic and the p threshold do not correspond under the declared test.
- p. 4 Results, Fatigue and desire to quit
F(2, 64) = 4.71
Only the authors can say what is right here.
Judge's note: For F with 2 and 64 df, p equals (1+2F/64)^(-32). With F = 4.71 this gives p of about .012, which is above .01. The .01 critical value is about 4.95. The neighbouring F(2, 64) = 5.13 does give p < .01. The paper declares no one-sided test or other adjustment. This is not an abstract-level result.
Second check: another model agreed
Report a problem with this finding
Wording Exp. 2 manipulation checks over the three speech conditions are printed with error df 31 ("F(2, 31) = 15.46", "F(2, 31) = 11.53"). The cell sizes implied by N = 39 and footnote 3 (t(18), t(18), t(17)) give about 29 participants in those three cells (error df about 26).
- p. 6 Exp. 2 Results, Manipulation check
F(2, 31) = 15.46, p < .001
- p. 6 Exp. 2 Results, second manipulation item
F(2, 31) = 11.53, p < .001
- p. 6 Footnote 3
t(17) = 1.90, p = .07
- p. 5 Exp 2 Participants
Participants were 39 undergraduate psychology stu-
Only the authors can say what is right here.
Judge's note: Page 5 gives N = 39 across four cells. Page 6 says the checks were asked "except in the control condition", so three groups. Footnote 3 gives t(18) for proattitudinal-choice vs control, t(18) for counterattitudinal-choice vs control, and t(17) for no-choice vs counterattitudinal-choice. I recomputed these from the Table 2 attempt means and SDs with cells of 10, 10, 9 and 10 (pro, counter-choice, no-choice, control). This gives t = 2.95, 1.71 and 1.90, matching the printed 2.94, 1.71 and 1.90. A split of 15/15/4/5, which would give 34 non-control participants, gives t of about 2.74 for pro vs control instead. So the three speech cells hold about 29 people and the error df should be about 26, not 31. The p < .001 statements are unaffected either way. Only the authors can say whether the df is a typo or reflects a different analysis set.
Second check: another model agreed
Report a problem with this finding
Wording Exp. 3 comedy-vs-sad film rating is reported as "t(29) = 4.62, p < .001" with N = 30, although a between-group comparison of 30 people allows at most 28 df (the other Exp. 3 tests use t(28)).
- p. 8 Exp. 3 Results, film perception
t(29) = 4.62, p < .001
- p. 7 Exp. 3 Method, Participants
Participants were 30 (11 male and 19 female)
- p. 7 Exp 3 Results
t(28) = 2.88, p < .01
- p. 8
than the sad video clip (M = 4.54), t(29) = 4.62, p < .001
Only the authors can say what is right here.
Judge's note: Page 7 says "Participants were 30", randomly assigned, and half of each condition saw each film, so film is between-subjects. The manipulation-check and anagram tests on pages 7 and 8 all use t(28). On page 8 the film-rating test is printed as t(29). A two-group test on 30 people allows at most 28 df, and no one-sample or paired test is described. The p value is unaffected. This is most likely a typo, but only the authors can confirm it.
Second check: another model agreed
Report a problem with this finding
Wording The General Discussion says "in Experiment 3, we found identical effects regardless of whether the person was trying to stifle a positive or a negative emotion", but the Results report only "There was no effect for type of movie".
- p. 11 General Discussion, Alternative Explanations
Moreover, in Experiment 3, we found identical effects regardless of whether the person was trying to stifle a positive or a negative emotion.
- p. 8 Experiment 3 Results, Anagram performance
There was no effect for type of movie.
Only the authors can say what is right here.
Judge's note: Table 3 gives only the suppress and no-regulation means. No means by film type and no suppression x film interaction test are reported, so "identical" claims more than the reported data show.
Second check: another model agreed
Report a problem with this finding
Wording Exp. 4 tiredness comparison between depletion and no-depletion conditions is printed as "t(83) = 2.79, p < .01" with N = 84, although a two-group comparison of 84 people allows at most 82 df (the ANOVA uses F(1, 80)).
- p. 9 Exp. 4 Results, Manipulation check
t(83) = 2.79, p < .01
- p. 9 Exp. 4 Method
Eighty-four undergraduate students
- p. 9 Results, Manipulation check
t(63) = 2.30
- p. 9 Results, tiredness
t(83)
Only the authors can say what is right here.
Judge's note: On page 9 the Method gives 84 participants. The ANOVA error df of 80 fits 84 people in four cells. The tiredness result is described as a between-condition comparison, which allows at most 82 df, yet t(83) is printed. A between-group comparison cannot have more than 82 df even if some data were missing. The concentration check's t(63) is below the maximum, so missing responses could explain it, and I do not count it as a defect. The p < .01 statement would hold with either df.
Second check: another model agreed
Report a problem with this finding
Wording Exp. 4 Discussion states "In the present study, ego depletion mediated the passive-option effect," but the reported analysis is a depletion × button-position interaction (moderation), not a mediation analysis.
- p. 10 Exp. 4 Discussion
ego depletion mediated the passive-option effect
- p. 10 Experiment 4 Discussion, first paragraph
In the present study, ego depletion mediated the passive-option effect.
Judge's note: Page 10 prints this sentence. The next paragraphs and Table 4 (differences of −14 and 54 s), together with the F(1, 80) = 5.64 interaction on page 9, describe an effect present only under depletion, which is moderation. No mediation analysis is reported. The word may be meant loosely, but as printed it misdescribes the statistical relation tested.
Second check: another model agreed
Report a problem with this finding
Typo Author note reads "supported by National Institute of Health Grants MH-51482 and MH-57039"; the agency name appears to be missing the plural "Institutes".
- p. 1 Author note
This research was supported by National Institute of Health Grants
Only the authors can say what is right here.
Judge's note: Rests on outside knowledge: the agency's official name is National Institutes of Health, and MH-prefixed grants come from NIMH within it. The crop of page 1 confirms the printed wording.
Second check: none (judge only)
Report a problem with this finding
Typo The Experiment 1 cover story uses two terms for one construct: "a follow-up measure for sensation memory" and later "to allow the sensory memory of the food to fade".
- p. 3 Experiment 1 Method, Procedure, cover story paragraph
follow-up measure for sensation memory
- p. 3 Experiment 1 Method, Procedure, next-but-one paragraph
allow the sensory memory of the
Judge's note: Both phrases are visible on page 3 and refer to the same cover-story construct. Experiment 3 also uses "sensory memory". This is a minor terminology inconsistency.
Second check: none (judge only)
Report a problem with this finding
Typo Footnote 4 contains the misspelling "most participants prefered to complete the procedure"; it should read "preferred".
- p. 9 Footnote 4
most participants prefered to complete the procedure
Judge's note: The misspelling is visible in the page 9 image, so it is not an extraction artefact.
Second check: none (judge only)