Results and Analyses -------------------- Participants We recruited participants (N = 111, males = 29, females = 82, M age = 20.54 years, SD = 2.15) from an undergraduate participant pool at University of Grenoble-Alpes. The participants were enrolled in a study on word, number recognition, and reaction time. The first 31 participants received course credit and the rest of the sample received 10 euros for their participation. Participants (n = 2) were excluded from the final analysis because they did not follow the instructions (n = 1), did not meet the meet the specified inclusion criteria for first-spoken language (n = 1), and age (18-30 years). In addition, 33 participants were excluded because their performance on the letter ‘e’ task or the MSIT fell below 80% accuracy or had mean reaction time or mean reaction time variability values that fell outside two standard deviations of the sample mean on the MSIT. The final sample comprised 32 participants in the hard letter ‘e’ (ego-depletion) condition and 46 participants in the easy letter ‘e’ (control) condition. Cédric Batailler and Camille Piollet served as the experimenter(s), and were blind to condition assignment. For those experimenters who were initially blind, blinding was done by giving a different and unrevealing name to both files (“easy” and “hard”). Our procedures followed the approved protocol and did not deviate from our preregistered plan, except that we recruited a little more than the planned 100 participants because more registered for our study. Critical analyses 1) Independent samples t-test comparing the ex-Gaussian fitted mean overall response time variability (RTV) for the incongruent items on the MSIT [note this is the ExGauss.I.RTVar.MSIT column in the output file] across the ego-depletion and control conditions. Ego-depletion: n = 32; M RTV= 0.3438; SD = 0.067; SE = 0.0119 Control: n = 46; M RTV = 0.3812; SD = 0.074; SE = 0.011 t(76) = -2.25, p = .027, d = .52 2) Independent samples t-test comparing the mean overall response time (RT) for the incongruent items on the MSIT [note this is the I_1_MeanRT.MSIT column in the output file] across the ego-depletion and control conditions. Ego-depletion: n = 32; M RT= 1.0365; SD = 0.1309; SE = 0.023 Control: n = 46; M RT = 1.0887; SD = 0.1381; SE = 0.0203 t(76) = -1.678, p = .095, d = .38 Although the difference between the “easy” and the “hard” condition was significant, mean response time variability and mean overall response time for the incongruent items on the MIST were higher in the “easy” (i.e. control) condition, we observed a mean difference opposite to our prediction. Since we were puzzled by these results, we checked coding of both “e” letter tasks, instructions in the “easy” and “hard” version. Everything was coded correctly. Moreover, scores on the fatigue, effort, difficulty and frustration scales are coherent with the “easy” and “hard” condition. Participants reported the “hard” task being more difficult, and requiring more effort (these difference were statistically significant) and also reported more fatigue and being more frustrated (with no significant differences). 3) A series of independent samples t-tests comparing participants’ mean ratings of effort, fatigue, and difficulty across the ego-depletion and control conditions (with positive t’s indicating larger rating in the ego-depletion group). Ego-depletion: Effort, M = 5.34; SD = 1.12; SE = 0.19 Fatigue, M = 4; SD = 1.74; SE = 0.30 Difficulty, M = 4.84; SD = 1.27; SE = 0.22 Frustration, M = 3.75; SD = 1.74; SE = 0.30 Control: Effort, M = 2.93; SD = 1.48; SE = 0.21 Fatigue, M = 3.63; SD = 1.86; SE = 0.27 Difficulty, M = 1.76; SD = 0.76; SE = 0.11 Frustration, M = 3; SD = 1.93; SE = 0.28 t-tests: Effort, t(76) = 7.92, M difference = -2.4, p < .001; Cohen's d = 1.8 Fatigue, t(76) = 0.88, M difference = -0.37, p = .38, Cohen's d = 0.2 Difficulty, t(76) = 13.34, M difference = -3.08, p < .001, Cohen's d = 3.1 Frustration, t(76) = 1.75, M difference = -0.75, p = .083, Cohen's d = 0.4 Supplemental analyses An independent samples t-test for differences in overall accuracy on the letter ‘e’ task [note this is the ‘Acc.Overall.LetE’ column in the output file] across the hard (ego-depletion) and easy (control) conditions: Ego-depletion: n = 32; M accuracy = 0.90; SD = 0.045; SE = 0.008 Control: n = 46; M accuracy = 0.99; SD = 0.005; SE = 0.0007 t(76) = 13.87, p < .001, d = 3.24