Modulating task outcome value to mitigate real-world procrastination via noninvasive brain stimulation

Reviewer #4 (Public review):

Summary:

The current study tested the effects of repeated sessions of tDCS targeting the DLPFC on procrastination behavior. The main outcome is that anodal versus sham DLPFC tDCS reduces procrastination behavior on both a short-term and a long-term scale up to six months after the stimulation sessions.

Strengths:

The current study tests competing models of procrastination with state-of-the-art high-definition transcranial electric stimulation. The study assesses stimulation effects on procrastination on both a short-term and a long-term scale, suggesting that repeated stimulation of the prefrontal cortex reduces procrastination on a time scale of up to six months.

Comments on revised version.

The manuscript has already been reviewed and revised before, and it seems that the quality of the manuscript has substantially improved as a result of this revision process. I agree with the other reviewers that one must be cautious with drawing conclusions regarding the cognitive mechanisms underlying this effect, as many different cognitive functions are implemented by the DLPFC. The effect sizes are surprisingly large, but I am satisfied with the reasons provided by the authors for the large effect sizes.

The authors successfully addressed my previous concerns on the manuscript.

Author response:

The following is the authors’ response to the previous reviews.

Public Reviews:

Reviewer #1 (Public review)

Summary:

The authors report the results of a tDCS brain stimulation study (verum vs sham stimulation of left DLPFC; between-subjects) in 46 participants, using an intense stimulation protocol over 2 weeks, combined with an experience-sampling approach, plus follow-up measures after 6 months.

Strengths:

The authors are studying a relevant and interesting research question using an intriguing design, following participants quite intensely over time and even at a follow-up time point. The use of an experience-sampling approach is another strength of the work.

Comments on revised version.

With the last round of revisions, the authors have now addressed my concerns.

Thank you to re-review this revision, and we all appreciate you kindly contributing to substantially improve the conceptualization, statistics and statements for this manuscript.

Reviewer #4 (Public review):

Summary:

The current study tested the effects of repeated sessions of tDCS targeting the DLPFC on procrastination behavior. The main outcome is that anodal versus sham DLPFC tDCS reduces procrastination behavior on both a short-term and a long-term scale up to six months after the stimulation sessions.

Strengths:

The current study tests competing models of procrastination with state-of-the-art high-definition transcranial electric stimulation. The study assesses stimulation effects on procrastination on both a short-term and a long-term scale, suggesting that repeated stimulation of the prefrontal cortex reduces procrastination on a time scale of up to six months.

Weaknesses:

The manuscript has already been reviewed and revised before, and it seems that the quality of the manuscript has substantially improved as a result of this revision process. I agree with the other reviewers that one must be cautious with drawing conclusions regarding the cognitive mechanisms underlying this effect, as many different cognitive functions are implemented by the DLPFC.

We do appreciate you to take valuable time offering those insightful and helpful comments on this revised manuscript. As you kindly raised, this revision has redrawn conclusions and statements on domain-specific mechanistic roles of DLPFC in interpreting why this neuromodulation treatments are effective.

One aspect of the current results that puzzles me is the strength of the current stimulation effects. Meta-analyses suggest that tDCS shows only small-to-moderate effect sizes (with Cohen's d around 0.5). While the authors report no effect sizes for their statistical models, the small p values, in combination with the unusually small sample size of 18 participants per group, suggests that the effect size must be rather large. Can the authors provide an estimate of the effect size of their stimulation effects? If they are considerably larger than to be expected, could the authors give an explanation for why their stimulation setup is showing much stronger effects than comparable high-definition tDCS studies on cognition or decision making?

Thank you for raising this very crucial question in effect size determination. We fully understand that this large effect size makes you puzzled, and that the limited sample size indeed attenuates detectable power in the statistics. We completely agree that reporting the actual effect sizes is essential for interpreting the magnitude of our findings, and we appreciate the opportunity to clarify why the observed effects in our study appear substantially larger than the small-to-moderate effect sizes typically reported in meta-analyses of tDCS studies on cognition and decision-making.

Following your suggestion, we have calculated the effect sizes for our primary outcomes. Based on the simple effect analyses (pre- vs. post-neuromodulation within the active neuromodulation group), we computed Cohen’s d for the within-group changes: for task-execution willingness, d = 2.37 (95% CI [1.49, 3.25]); for the actual procrastination rate, d = 1.52 (95% CI [0.87, 2.16]).

We acknowledge that these effect sizes are considerably larger than the typical d ≈ 0.5 reported in the tDCS literature. After careful consideration, we attribute this discrepancy to three methodological and conceptual differences between our study and typical cognitive/decision-making tDCS studies. First, the small-to-moderate effect sizes are observed in studies utilizing a single-session tDCS protocol, yet our study employed an intensive 7-session HD-tDCS protocol over 15 days. Therefore, multi-session tDCS that induces cumulative, activity-dependent long-term potentiation (LTP)-like plasticity may substantially amplify and consolidates behavioral effects compared to single-session stimulation (Ke et al., 2023; Zhong et al., 2021). Second, most tDCS studies on cognition recruit healthy young adults who often perform near ceiling on laboratory tasks, leaving little "room for improvement" and thereby constraining the observable effect size. In our study, we strictly screened for severe chronic procrastinators. Because our participants had severe baseline deficits in task execution, the "ceiling space" for behavioral improvement was much larger, naturally inflating the observable effect size of the intervention. Lastly, given all the procrastinators completed tasks in the last session (0% procrastination rate in the active group, without within-group variance), mathematically, this boundary variable (0% vs 100%) artificially inflates the effect size estimate when calculating Cohen’s d with a near-zero post-test standard deviation.

Nevertheless, as a sensitivity analysis, those findings are confirmed by Beta regression model addressing the risks of boundary variables, indicating that the potential inflation of effect sizes is statistically acceptable:

In summary, while the observed Cohen's d values are unusually large, they are contextually justified by the cumulative nature of our multi-session protocol, the targeted clinical-like population, and the mathematical properties of bounded behavioral metrics.

Results Section (Page 9, Line 441-444)

“... For procrastination willingness, results showed a statistically significant interaction effect between multi-session neuromodulations and groups (β = -7.84, SE = 1.80, t = -4.36, DF = 45.6, p < .001, Cohen d = 2.37, 95% CI: 1.49-3.25; Fig. 3A and Fig. S2a).”

Results Section (Page 9, Line 454-458)

“... Similarly, a statistically significant interaction effect was identified here (β = -7.37, SE = 2.40, t = -3.02, DF = 46.6, p = .004, Cohen d = 1.52, 95% CI: 0.87-2.16), and the simple effect analysis further revealed decreased actual procrastination rates after ms-tDCS in the active neuromodulation group.”

Regarding the strengths of the stimulation effects, I moreover found remarkable that the post-test procrastination rate was 100% in all (!) participants in the DLPFC group (figure 3F). I admit that it is hard to trust results that have no individual variation at all. This means that all participants are perfect responders to tDCS, which is again at variance what one typically expects for tDCS (where one usually has many non-responders). Do the authors have an explanation for this?

Thank you for raising this highly important and helpful comment. Indeed, we fully understand that this result (a 100% task complete rate among all participants in the DLPFC group) is somewhat extraordinary. This pattern was equally striking to us when unblinded the data. After carefully scrutinizing the data and statistics, we are thrilled to confirm that this pattern is true. In support of this observation, we were gratified to receive numerous thank-you letters from participants who engaged in active neuromodulation. They expressed gratitude to us, and reported that they have substantially ameliorated procrastination behavior in real-life activities after completing the trial. While this does not constitute formal scientific evidence, we are also glad to see the benefits of this neuromodulation for those procrastinators.

Two reasons could account for this pattern herein. One interpretation is to attribute this pattern to “floor effect”. In the present study, the procrastination rate was calculated as 1 minus the task-completion rate (e.g., 80%, 60%, 40%) by the deadline. At last stimulation sessions (#6 and #7), all the participants completed their real-life tasks before the deadline, yielding a 0% (1 minus 100% completion rate) procrastination rate, without any between-individual variation. Thus, rather than there being no individual variation in procrastination, this scalar – the procrastination rate - is too insensitive to capture subtle differences per se. For instance, although participants #1 and #2 both showed a 0% procrastination rate - meaning that both completed their tasks before the deadline - Participant #1 might have completed it 3 hours before the deadline, whereas Participant #2 might have completed it only 10 minutes before. In this case, the “scalar inflation” emerges to let us perceive that both participants have equivalent procrastination rates, although participant #2 may have a higher procrastination level than #1. As conceptually defined in the field, procrastination is contextualized as “not completing a task before the deadline”. Thus, if this task is completed before the deadline, regardless of whether it was finished close to or far in advance of the deadline, this case is defined as “no procrastination”. In the present study, the primary outcome is whether a participant procrastinated on a real-life task before the deadline in real-world settings, irrespective of when she/he completed this task. Thus, this scalar - procrastination rate - fits our conceptualization of procrastination.

Another reason is the potential accumulative effects from sequential multi-session tDCS stimulation, as we explained above. As shown in Mann-Kendall trend tests, the procrastination rates show a significant linear downtrend in the active neuromodulation group across sessions, even after removing sessions #6 and #7. This indicates that the improvements of going against procrastination may be sequentially accumulative along with the increase in sessions, implying a potential “dose-dependent effect”. Despite a speculative interpretation, this “dose-dependent effect” in neuromodulation has been well-documented in previous studies, showing the robustly linear association between the number of sessions and effectiveness (c.f., Cole et al., 2020; Hutton et al., 2023; Sabé et al., 2024; Schulze et al., 2018). Therefore, although this extreme pattern is somewhat extraordinary compared to previous observations, it makes sense.

We also conducted robustness check by removing sessions #6, #7, and both, to validate whether this results were biased by “scalar inflation”. We do believe that this analysis could support statistical robustness to go against potential biases from extreme cells. By doing so, we found that all the group*treatment_day interaction effects remained significant when removing either session #6 or session #7 (or even both, all p-values < .05), indicating high statistical robustness. Please see Table S3 and Table S4.

Taken together, in spite of their being extraordinary, we confirm that those findings are statistically robust to extreme outliers. As you kindly suggested, we have added those findings of the robustness check into the revised Supplemental Materials section.

In any case, I am surprised by the rather small sample size. Due to the small effect sizes for tDCS, it is common to have a minimum of 30 subjects per group in between-subject designs. According to G*Power, a between-subject design with 17 subjects per group could detect only relatively large effect sizes of Cohen's d = 0.99 (alpha = 5%, power = 80%, independent-samples t-test). As explained above, this is far above the effect size that can be expected for tDCS. In addition, small samples bear the risk that results strongly depend on outliers in the data, which might explain the strong effect size observed in the current study. The small sample size should be discussed as a major limitation of the current study and that the results need to be replicated by studies with larger sample sizes. Moreover, to rule out that the results are driven by outlier in the data, the authors should show individual data points in all plots showing empirical data.

We sincerely thank you for this highly constructive and methodologically sound critique. We completely agree that sample size is a critical consideration in tDCS research, and that visualizing individual data points is essential to rule out the possibility that our findings are driven by outliers.

We acknowledge that our sample size is smaller than the ~30 per group often recommended for detecting small-to-moderate effects in general cognitive tDCS meta-analyses. We have determined this a priori effect size based on the existing work we published previously (Xu et al., 2023, J Exp Psychol Gen;152(4):1122-1133). In our pilot study (Xu et al., 2023), we identified a significant interaction effect between the single-session tDCS stimulation (active vs sham) and time (pre-test vs post-test) (t = 2.38, p = .02, n = 27; 95% CI [0.14, 1.49]) for changing procrastination willingness in the laboratory settings, indicating a medium effect size. Based on this specific empirical foundation, GPower indicated that a total sample size of 34 (17 per group) was sufficient to achieve 80% power (please see GPower output below). To account for potential attrition, we aimed to recruit 36 participants (18 per group), ultimately retaining 46 participants (23 per group) after exclusions. While we stand by this a priori justification, we fully agree with your overarching point that this remains a constraint.

We completely agree with your observation regarding the plots. The apparent absence of data points in the previous versions of Figures 3B and 3F was not due to data exclusion, but rather to severe overplotting. Because multiple participants in the active neuromodulation group achieved identical scores (e.g., 0% procrastination rate or 100% task-execution willingness in later sessions), their data points perfectly overlapped, making it appear as though only ~10 points were present. As you helpfully suggested, we now employ jittered scatter plots with adjusted transparency, ensuring that all 23 individual data points per group are clearly visible, even when values are identical. As these revised figures demonstrate, the significant group differences reflect a consistent, cohort-wide shift rather than the influence of isolated outliers. Those

As you rightly suggested, we have explicitly framed the small sample size as a major limitation and emphasized the necessity for large-scale replication. We have strengthened the wording in the Limitations section to explicitly mention the risk of outlier dependency and the need for larger cohorts.

Legend Section (Page 28, Line 1161-1163)

“… To ensure transparency and rule out outlier-driven effects, individual data points for all participants (N=23 per group) are overlaid on the bars using a jittered distribution to prevent overplotting of identical values.”

Discussion Section (Page 13, Line 691-696)

“… a major limitation of the current study is the relatively small sample size (total N = 46). While this was determined a priori based on our specific pilot study, small samples inherently bear a higher risk of being influenced by outliers and may overestimate effect sizes compared to large-scale meta-analytic expectations for tDCS. Therefore, these findings warrant caution in generalization and necessitate rigorous replication in larger, adequately powered cohorts.”

Related to this, in the figure showing individual data points (3B/F), I count only around 10 data points per tDCS group for the 18 participants per group. I ask the authors to modify the plot that the data points from all participants can be seen (for example, by adding some noise on the x-axis for participants with the same value on the y axis).

Thank you for this kind reminder. As we replied above, those plots have been redrawn by adding the jitters, which favor the readability as you kindly suggested.

Another surprising aspect of the data is that repeated sessions of tDCS change procrastination behavior up to six months after stimulation. Do the authors think that their tDCS setup leads to such long-lasting neuroplastic changes, and if yes, can they cite prior work where similar dosages of tDCS also showed such long-lasting effects? Or could the results be explained by learning effects, for example because participants in the DLPFC group learned during the repeated tDCS sessions that it feels internally rewarding to finish one's tasks instead of procrastinating them, and they still benefit from this kind of "learned industriousness" 6 months later? In any case, in my view it is important to be more specific about how seven sessions of tDCS can affect behavior half a year later.

We sincerely thank the reviewer for this highly insightful and thought-provoking comment. The concept of "learned industriousness" is particularly apt and captures a crucial alternative mechanism that we must address. We agree that explaining how seven sessions of tDCS can affect behavior half a year later requires a nuanced discussion of both neurobiological and behavioral learning mechanisms.

Regarding the first point, we do believe that our multi-session protocol can induce long-lasting neuroplastic changes. While single-session tDCS effects are typically transient, cumulative neurobiological evidence demonstrates that repeated, multi-session protocols (typically ranging from 5 to 10 sessions) can induce activity-dependent, long-term potentiation (LTP)-like plasticity that consolidates over time (Agboada et al., 2020; Au et al., 2017; Jannati et al., 2023). Our 7-session protocol falls squarely within this range of "intensified dosing" designed to promote such consolidation. Meta-analyses and empirical studies on multi-session tDCS have shown that such protocols can produce behavioral and neurophysiological effects lasting weeks to months, particularly when targeting prefrontal regions involved in value-based decision-making and cognitive control (e.g., Brunoni et al., 2013; Sabé et al., 2024; Woodham et al., 2025).

Furthermore, we completely agree with you for this alternative explanation regarding learning effects. It is highly plausible that participants in the active group, experiencing reduced task aversiveness and increased outcome value during the intervention, learned that completing tasks is internally rewarding. This aligns perfectly with the psychological concept of "learned industriousness" (Eisenberger, 1992), where the reinforcement of effortful behavior makes future engagement more likely. We explicitly acknowledge that repeated exposure to the experience-sampling protocol and the positive feedback of task completion could facilitate this kind of behavioral learning. More importantly, we argue that the learning effects are not bad things in this neuromodulation, and the learning effect and the neuroplasticity may be synergistic. The sham control group underwent the exact same experience-sampling protocol, reported real-life tasks, and had the identical opportunity for "learned industriousness" through feedback. However, as identified in the half-year follow-up, the sham group did not exhibit the same progressive improvement during the intervention, nor did they sustain a significant reduction in procrastination at the 6-month follow-up (their rates returned to near-baseline levels). This divergence suggests that while learning may play a role, the active neuromodulation likely provided the necessary neuroplastic "boost" (e.g., by enhancing prefrontal value-encoding circuits) that facilitated, accelerated, and consolidated this learning, making the behavioral change durable. Without the neuromodulatory enhancement, the mere exposure to the protocol was insufficient to produce long-term change.

As you kindly suggested, we have explicitly incorporated this nuanced discussion into the revised manuscript, by citing relevant literature on multi-session tDCS plasticity, explicitly acknowledging the "learned industriousness" hypothesis, and reiterating the limitation of having only a single follow-up point.

Discussion Section (Page 12, Line 627-634)

“... Despite statistically supporting the TDM, we acknowledge that alternative neurocognitive mechanisms could contribute to the observed reductions in procrastination. For instance, repeated exposure to the experience-sampling protocol may have enhanced participants’ awareness of task progress or facilitated feedback-based learning, thereby increasing the subjective value of goal completion independent of DLPFC neuromodulation. Participants in the active group may have learned during the repeated sessions that completing tasks feels internally rewarding, thereby benefiting from a form of “learned industriousness” (Eisenberger, 1992) that persists months later.”

Discussion Section (Page 14, Line 719-723)

“... we explicitly note that a single 6-month follow-up timepoint cannot definitively establish the stability or trajectory of these effects. Future studies incorporating multiple longitudinal assessments (e.g., 1-month, 3-month, 6-month, 12-month) are required to substantiate claims about long-term retention and to disentangle the precise contributions of neuroplasticity versus behavioral learning.”

Lastly, the link to the data repository works, but I could not inspect the data because I was asked to request access to the data, which I did not do in order to remain anonymous.

Thank you a lot to take invaluable to review our data and code in this repository. As we reported previously, all the data and code to support those findings have been deposited in the eLife online submission system for your reviews and scrutiny before this manuscript is formally published. As the editorial policy of eLife on VOR (Version of Record) instructed, to prevent from mixture of codes and data across multiple round of revisions, those data and codes in the final version would be released once this paper is formally published. Please do not worry for the anonymity policy. This is a public peer review, and it thus enables those helpful comments that you kindly suggested to be public when this manuscript is formally published. Again, thank you to substantially contribute on this revised manuscript by sharing those helpful suggestions.

Recommendations for the authors:

Editors note: We encourage the authors to consider the remaining reviewer concerns and revise the manuscript accordingly.

Thank you so much for this warm and kind reminder. We have addressed all of those concerns that remained by the new Reviewer #4, point-by-point. All the co-authors do appreciate you for handling our manuscript, and for contributing those fruitful and helpful comments. We do believe that the quality of this manuscript has been substantially improved, benefiting from this editorial process.

References

Agboada, D., Mosayebi-Samani, M., Kuo, M. F., & Nitsche, M. A. (2020). Induction of long-term potentiation-like plasticity in the primary motor cortex with repeated anodal transcranial direct current stimulation - Better effects with intensified protocols? Brain Stimulation, 13(4), 987–997. https://doi.org/10.1016/j.brs.2020.04.009

Au, J., Karsten, C., Buschkuehl, M., & Jaeggi, S. M. (2017). Optimizing transcranial direct current stimulation protocols to promote long-term learning. Journal of Cognitive Enhancement, 1(1), 65–72. https://doi.org/10.1007/s41465-017-0007-6

Brunoni, A. R., Boggio, P. S., Ferrucci, R., Priori, A., & Fregni, F. (2013). Transcranial direct current stimulation: challenges, opportunities, and impact on psychiatry and neurorehabilitation. Frontiers in Psychiatry, 4, 19. https://doi.org/10.3389/fpsyt.2013.00019

Cole, E. J., Stimpson, K. H., Bentzley, B. S., Gulser, M., Cherian, K., Tischler, C., Nejad, R., Pankow, H., Choi, E., Aaron, H., Espil, F. M., Pannu, J., Xiao, X., Duvio, D., Solvason, H. B., Hawkins, J., Guerra, A., Jo, B., Raj, K. S., Phillips, A. L., … Williams, N. R. (2020). Stanford accelerated intelligent neuromodulation therapy for treatment-resistant depression. The American Journal of Psychiatry, 177(8), 716–726. https://doi.org/10.1176/appi.ajp.2019.19070720

Eisenberger, R. (1992). Learned industriousness. Psychological Review, 99(2), 248–267. https://doi.org/10.1037/0033-295X.99.2.248

Hutton, T. M., Aaronson, S. T., Carpenter, L. L., Pages, K., Krantz, D., Lucas, L., Chen, B., & Sackeim, H. A. (2023). Dosing transcranial magnetic stimulation in major depressive disorder: Relations between number of treatment sessions and effectiveness in a large patient registry. Brain Stimulation, 16(5), 1510–1521. https://doi.org/10.1016/j.brs.2023.10.001

Jannati, A., Oberman, L. M., Rotenberg, A., & Pascual-Leone, A. (2023). Assessing the mechanisms of brain plasticity by transcranial magnetic stimulation. Neuropsychopharmacology, 48(1), 191–208. https://doi.org/10.1038/s41386-022-01453-8

Ke, Y., Liu, S., Chen, L., et al. (2023). Lasting enhancements in neural efficiency by multi-session transcranial direct current stimulation during working memory training. npj Science of Learning, 8(1), Article 23. https://doi.org/10.1038/s41539-023-00200-y

Sabé, M., Hyde, J., Cramer, C., Eberhard, A., Crippa, A., Brunoni, A. R., Aleman, A., Kaiser, S., Baldwin, D. S., Garner, M., Sentissi, O., Fiedorowicz, J. G., Brandt, V., Cortese, S., & Solmi, M. (2024). Transcranial magnetic stimulation and transcranial direct current stimulation across mental disorders: A systematic review and dose-response meta-analysis. JAMA Network Open, 7(5), e2412616. https://doi.org/10.1001/jamanetworkopen.2024.12616

Schulze, L., Feffer, K., Lozano, C., Giacobbe, P., Daskalakis, Z. J., Blumberger, D. M., & Downar, J. (2018). Number of pulses or number of sessions? An open-label study of trajectories of improvement for once- vs. twice-daily dorsomedial prefrontal rTMS in major depression. Brain Stimulation, 11(2), 327–336. https://doi.org/10.1016/j.brs.2017.11.002

Woodham, R. D., Selvaraj, S., Lajmi, N., Hobday, H., Sheehan, G., Ghazi-Noori, A.-R., Lagerberg, P. J., Rizvi, M., Kwon, S. S., Orhii, P., Maislin, D., Hernandez, L., Machado-Vieira, R., Soares, J. C., Young, A. H., & Fu, C. H. Y. (2025). Home-based transcranial direct current stimulation treatment for major depressive disorder: a fully remote phase 2 randomized sham-controlled trial. Nature Medicine, 31(1), 87–95. https://doi.org/10.1038/s41591-024-03305-y

Zhong, M., Cywiak, C., Metto, A. C., Liu, X., Qian, C., et al. (2021). Multi-session delivery of synchronous rTMS and sensory stimulation induces long-term plasticity. Brain Stimulation, 14(4), 884–894. https://doi.org/10.1016/j.brs.2021.05.003