REVIEW 3 major objections 1 minor 65 references
Generative AI game Reverie reduces stress levels in a 14-day student pilot.
Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →
T0 review · grok-4.3
2026-05-25 06:32 UTC pith:JPTMFSUZ
load-bearing objection The paper builds a ChatGPT-powered Unity game for stress relief and runs a 14-day pilot on 20 people, but the single-arm design does not support the causal claim in the title. the 3 major comments →
A Generative AI Driven Interactive Narrative Serious Game for Stress Relief and Its Randomized Controlled Pilot Study
The pith
A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.
Core claim
The paper claims that the Gen-AI driven serious game Reverie, built with Unity and ChatGPT, demonstrates feasibility and preliminary efficacy for stress relief. Participants showed a significant reduction in stress levels (p=.016) over the intervention period, with excellent user experience and associated improvements in cognitive emotion regulation strategies. It proposes a Gen-AI driven design framework for serious games for stress relief in a personalized digital intervention context.
What carries the argument
Reverie, the interactive narrative serious game that combines the Unity engine with ChatGPT to deliver generative, personalized gameplay for stress relief.
Load-bearing premise
The observed stress reductions result from playing the game rather than from the passage of time, participant expectations, or other unmeasured factors.
What would settle it
A follow-up study in which a no-game control group shows the same stress decline over 14 days as the Reverie group.
If this is right
- Generative AI can be integrated into game engines to create personalized narrative interventions for stress.
- Continued use of such a game over multiple days can produce cumulative stress reduction.
- The intervention is associated with shifts toward more adaptive cognitive emotion regulation strategies.
- A design framework combining LLMs and serious games offers a route to scalable digital mental health tools.
Where Pith is reading between the lines
- Larger randomized trials with an explicit control arm would be needed to isolate the game's causal contribution.
- The approach could be extended to track whether stress benefits persist after the 14-day period ends.
- Real-time AI adaptation to user-reported mood might further strengthen the personalization effect.
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The paper presents the development of 'Reverie', a generative AI-driven interactive narrative serious game built with Unity and ChatGPT for stress relief. It reports a 14-day pilot study with 20 students experiencing moderate to high stress, claiming that the game significantly reduced stress levels (p=.016), demonstrated excellent user experience, and was associated with improvements in cognitive emotion regulation strategies, while proposing a Gen-AI design framework for such interventions.
Significance. If the efficacy results were supported by appropriate controls and measures, the work would offer preliminary evidence for the feasibility of LLM-integrated serious games in mental health, along with a proposed design framework. The small sample and single-arm design limit the strength of any causal conclusions about stress relief.
major comments (3)
- [Title / Abstract] Title and Abstract: The title describes a 'Randomized Controlled Pilot Study', but the methods describe only a single-arm 14-day intervention on 20 participants with no reference to randomization procedures, a control condition, or between-group contrasts. This mismatch is load-bearing because the headline result attributes the p=.016 stress reduction to the game itself.
- [Abstract / Results] Abstract / Results: The claim that 'Reverie' significantly reduced stress levels (p=.016) is presented as evidence of the intervention's effect, yet the protocol is a within-subject pre-post design with no control arm. Without a control group, the p-value shows only change over time and cannot rule out regression to the mean, repeated assessment, or expectation effects.
- [Abstract / Methods] Abstract / Methods: No information is supplied on the stress measurement instruments, the specific statistical tests, handling of dropouts, inclusion/exclusion criteria, or the cognitive emotion regulation scales used, despite the small sample (n=20) and the efficacy claims.
minor comments (1)
- [Abstract] The abstract states that participants 'used Reverie as a stress-relief intervention' but provides no further protocol details such as session frequency or duration.
Simulated Author's Rebuttal
We thank the referee for these constructive comments, which correctly identify inaccuracies in the title, overstatements in the results interpretation, and insufficient methodological detail in the abstract. We will undertake major revisions to correct these issues while preserving the pilot nature of the work.
read point-by-point responses
-
Referee: [Title / Abstract] Title and Abstract: The title describes a 'Randomized Controlled Pilot Study', but the methods describe only a single-arm 14-day intervention on 20 participants with no reference to randomization procedures, a control condition, or between-group contrasts. This mismatch is load-bearing because the headline result attributes the p=.016 stress reduction to the game itself.
Authors: We agree that the title is inaccurate and misleading. The study is a single-arm pre-post pilot with no randomization or control arm. We will revise the title to 'A Generative AI Driven Interactive Narrative Serious Game for Stress Relief: A Pilot Study' and update the abstract to consistently describe the design as a single-arm intervention. revision: yes
-
Referee: [Abstract / Results] Abstract / Results: The claim that 'Reverie' significantly reduced stress levels (p=.016) is presented as evidence of the intervention's effect, yet the protocol is a within-subject pre-post design with no control arm. Without a control group, the p-value shows only change over time and cannot rule out regression to the mean, repeated assessment, or expectation effects.
Authors: We accept this critique. The current wording implies a causal effect that the design cannot support. We will rephrase the abstract and results to report that stress levels decreased significantly over the 14-day period (p=.016) within this single-arm design, and we will add explicit discussion of limitations including regression to the mean, repeated testing, and expectation effects. revision: yes
-
Referee: [Abstract / Methods] Abstract / Methods: No information is supplied on the stress measurement instruments, the specific statistical tests, handling of dropouts, inclusion/exclusion criteria, or the cognitive emotion regulation scales used, despite the small sample (n=20) and the efficacy claims.
Authors: We agree that the abstract is missing these essential details. Although the full methods section contains them, we will expand the abstract to specify the stress instrument (Perceived Stress Scale), statistical test (paired t-test or appropriate non-parametric equivalent), inclusion criteria (students reporting moderate to high stress), dropout handling, and the cognitive emotion regulation measure (CERQ). revision: yes
Circularity Check
No circularity: empirical pilot study with no derivations or fitted parameters
full rationale
The paper reports results from a 14-day pilot study on 20 participants, including a p=.016 stress reduction finding and user experience metrics. No equations, derivations, parameters, or self-citations form a load-bearing chain; claims rest directly on collected data and standard statistical reporting rather than any self-referential construction. This matches the default expectation of no significant circularity for non-theoretical empirical work.
Axiom & Free-Parameter Ledger
read the original abstract
Background: Stress has become a widespread phenomenon, and serious games are increasingly recognized as engaging tools for stress relief. However, despite the rapid advancement of Generative Artificial Intelligence (Gen-AI), its integration into stress-relief serious games remains insufficiently explored. Objective: This study aimed to address this gap by developing "Reverie", an Gen-AI driven serious game powered by the Unity engine and ChatGPT, and to preliminarily evaluate its effectiveness in stress reduction, user experience, and cognitive emotion regulation. Methods: A 14-day pilot study was conducted with 20 students experiencing moderate to high levels of stress. Participants used "Reverie" as a stress-relief intervention. Stress levels, user experience, and cognitive emotion regulation strategies were assessed to examine the game's feasibility and preliminary efficacy. Results: The results showed that "Reverie" significantly reduced participants' stress levels over the intervention period (p=.016*), indicating a cumulative positive effect. In addition, the game demonstrated excellent user experience and was associated with improvements in cognitive emotion regulation strategies. Conclusions: This study proposes a Gen-AI driven design framework for serious games for stress relief. Besides, this pilot study provides initial support for the feasibility and promise of combining LLM-driven gameplay in a personalized digital intervention context.
Reference graph
Works this paper leans on
-
[1]
F. Ghasemi, D.Q. Beversdorf, K.C. Herman, Stress and stress responses: a narrative literature review from physiological mechanisms to intervention approaches, J. Pac. Rim Psychol. 18 (2024) 18344909241289222
work page 2024
-
[2]
M. Mehta, M.M. Singh, S.K. Gupta, A. Kushal, Study of stress among health care professionals: a systemic review, Int. J. Res. Found. Hosp. Healthc. Adm. 6 (1) (2018) 6-11
work page 2018
-
[3]
I. Prilleltensky, M. Neff, A. Bessell, Teacher stress: what it is, why it's important, how it can be alleviated, Theory Pract. 55 (2) (2016) 104-111
work page 2016
-
[4]
Y. Amanvermez, M. Rahmadiana, E. Karyotaki, L. de Wit, D.D. Ebert, R.C. Kessler, et al., Stress management interventions for college students: a systematic review and meta-analysis, Clin. Psychol. Sci. Pract. 30 (4) (2023) 423
work page 2023
-
[5]
S.E. Williams, J. Cumming, G.M. Balanos, The use of imagery to manipulate challenge and threat appraisal states in athletes, J. Sport Exerc. Psychol. 32 (3) (2010) 339-358
work page 2010
- [6]
-
[7]
Chrousos, Stress and disorders of the stress system, Nat
G.P. Chrousos, Stress and disorders of the stress system, Nat. Rev. Endocrinol. 5 (7) (2009) 374-381
work page 2009
- [8]
-
[9]
A. Reichenberg, J.H. MacCabe, Feeling the pressure: work stress and mental health, Psychol. Med. 37 (8) (2007) 1073-1074
work page 2007
-
[10]
K.J. Reddy, K.R. Menon, A. Thattil, Academic stress and its sources among university students, Biomed. Pharmacol. J. 11 (1) (2018) 531-537
work page 2018
-
[11]
Aneshensel, Social stress: theory and research, Annu
C.S. Aneshensel, Social stress: theory and research, Annu. Rev. Sociol. 18 (1) (1992) 15-38
work page 1992
-
[12]
R.D. Conger, M.A. Rueter, G.H. Elder Jr., Couple resilience to economic pressure, J. Pers. Soc. Psychol. 76 (1) (1999) 54
work page 1999
- [13]
-
[14]
P. Nixon, D.D. Ebert, L. Boß, P. Angerer, N. Dragano, D. Lehr, The efficacy of a web-based stress management intervention for employees experiencing adverse working conditions and occupational self-efficacy as a mediator: randomized controlled trial, J. Med. Internet Res. 24 (10) (2022) e40488
work page 2022
- [15]
-
[16]
M. Cozzolino, D.R. Vivo, L. Girelli, P. Limone, G. Celia, The evaluation of a mind-body intervention (MBT-T) for stress reduction in academic settings: a pilot study, Behav. Sci. 10 (8) (2020) 124
work page 2020
-
[17]
N.D. Eneogu, C.K. Ugwuanyi, C.S. Ugwuanyi, Efficacy of cognitive behavioral therapy on academic stress among rural community secondary school economics students: a A generative AI driven interactive narrative serious game for stress relief and its randomized controlled pilot study *Email: tingchenhsu.ac@gmail.com; 13 randomized controlled evaluation, J. R...
work page 2024
-
[18]
K. Khurshid, R. Mushtaq, U. Rauf, N. Anwar, Q. Abbas, S. Aljhani, et al., Cognitive behavior therapy for academic burnout, procrastination, self-handicapping behavior, and test anxiety among adolescents: a randomized control trial, BMC Psychol. 13 (1) (2025) 94
work page 2025
-
[19]
T. Evriani, N.A. Fardana, How does the effectiveness of cognitive behavior therapy in reducing academic anxiety influence the academic procrastination of undergraduate students? Buana Pendidikan 20 (1) (2024) 22-28
work page 2024
-
[20]
Z. Li, R. Yao, S.H. Cho, Economic insecurity, perceived stress, and depressive symptoms: a longitudinal study on mental health, Appl. Res. Qual. Life 20 (4) (2025) 1611-1628
work page 2025
-
[21]
T. Richardson, A. Enrique, C. Earley, A. Adegoke, D. Hiscock, D. Richards, The acceptability and initial effectiveness of “Space From Money Worries”: an online cognitive behavioral therapy intervention to tackle the link between financial difficulties and poor mental health, Front. Public Health 10 (2022) 739381
work page 2022
-
[22]
L. Tang, Q.X. Liu, Y. He, H. Peng, M. Luo, H. Zheng, et al., Systematic interventions based on the stress-induced situation, affective, bodily, and cognitive reactions framework to mitigate psychological distress in lung cancer patients post-thoracoscopic surgery: a randomized clinical trial, Front. Psychol. 16 (2025) 1511622
work page 2025
-
[23]
J.M. Stagl, S.C. Lechner, C.S. Carver, L.C. Bouchard, L.M. Gudenkauf, D.R. Jutagir, et al., A randomized controlled trial of cognitive-behavioral stress management in breast cancer: survival and recurrence at 11-year follow-up, Breast Cancer Res. Treat. 154 (2) (2015) 319-328
work page 2015
-
[24]
F.C. Omeke, M.O. Ede, C.L. Chukwu, P.N. Aroh, C.O. Onyeanusi, T.O. Ozor, et al., Can REBT reduce academic stress and increase adjustment in rural students? J. Ration. Emot. Cogn. Behav. Ther. 42 (4) (2024) 699-721
work page 2024
-
[25]
A. Puolakanaho, R. Lappalainen, P. Lappalainen, J.S. Muotka, R. Hirvonen, K.M. Eklund, et al., Reducing stress and enhancing academic buoyancy among adolescents using a brief web-based program based on acceptance and commitment therapy: a randomized controlled trial, J. Youth Adolesc. 48 (2) (2019) 287-305
work page 2019
- [26]
-
[27]
J.I. Gallego-Gómez, S. Balanza, J. Leal-Llopis, J.A. García-Méndez, J. Oliva-Pérez, J. Doménech-Tortosa, et al., Effectiveness of music therapy and progressive muscle relaxation in reducing stress before exams and improving academic performance in nursing students: a randomized trial, Nurse Educ. Today 84 (2020) 104217
work page 2020
- [28]
-
[29]
L. Reinecke, Games and recovery: the use of video and computer games to recuperate from stress and strain, J. Media Psychol. 21 (3) (2009) 126-142
work page 2009
-
[30]
E. Usta, M. Inozu, The use of serious games in psychological interventions for anxiety disorders and related psychopathologies: a systematic review, Curr. Psychol. 43 (2024) 12610-12633
work page 2024
-
[31]
S. Lu, S. Ni, T. Bai, Promoting migrant adolescent resilience through a virtual reality videogame: a randomized trial, Res. Soc. Work Pract. (2025) 10497315251342210
work page 2025
-
[32]
J. Zhan, C. Liu, Z. Wang, Z. Cai, J. He, Effects of game-based digital interventions for mental disorders: a meta-analysis, J. Affect. Disord. 362 (2024) 731–741
work page 2024
-
[33]
M. Van Rooij, A. Lobel, O. Harris, N. Smit, I. Granic, DEEP: a biofeedback virtual reality game for children at-risk for anxiety, in: Proceedings of the 2016 CHI Conference Extended Abstracts on Human Factors in Computing Systems, 2016, pp. 1989-1997
work page 2016
-
[34]
M. Zeiler, S. Vögl, U. Prinz, N. Werner, G. Wagner, A. Karwautz, N. Zeller, L. Ackermann, K. Waldherr, Game Design, Effectiveness, and Implementation of Serious Games Promoting Mental Health Literacy in Children and Adolescents: Systematic Review, JMIR Mental Health 12 (2025) e67418
work page 2025
-
[35]
N.A. Mohamad Yahaya, D.R. Awang Rambli, S. Sulaiman, F. Merienne, E. Alyan, Design of game-based virtual forests for psychological stress therapy. Forests 14(2) (2023) 288
work page 2023
-
[36]
F. Pallavicini, E. Orena, L. Arnoldi, F. Achille, S. Stefanini, M. Cassa, et al., Effects and acceptability of a 1-week home-based virtual reality training for supporting the management of stress and anxiety: randomized pilot trial, JMIR Serious Games 13(1) (2025) e50326
work page 2025
-
[37]
Y.K. Heng, J.S.Y. Liew, M.F.I.L. Abdullah, Y. Tang, N. Prestopnik, ReWIND: a CBT-based serious game to improve cognitive emotion regulation and anxiety disorder, Int. J. Serious Games 10 (3) (2023) 43-65
work page 2023
-
[38]
T.C. Hsu, Z. Zhang, Z. Chen, Y. Liu, Y. Liu, Designing More Engaging Serious Games to Support Students’ Mental Health: A Pilot Study Based on A CBT-Informed Design Framework, arXiv preprint arXiv:2604.15662, 2026
work page internal anchor Pith review Pith/arXiv arXiv 2026
-
[39]
S.S. Sengar, A.B. Hasan, S. Kumar, F. Carroll, Generative artificial intelligence: a systematic review and applications, Multimed. Tools Appl. 84 (21) (2025) 23661-23700
work page 2025
- [40]
-
[41]
J. Ashkinaze, J. Mendelsohn, Q. Li, C. Budak, E. Gilbert, How AI ideas affect the creativity, diversity, and evolution of human ideas: evidence from a large, dynamic experiment, in: Proceedings of the ACM Collective Intelligence Conference, 2025, pp. 198-213
work page 2025
-
[42]
R. Deng, M. Jiang, X. Yu, Y. Lu, S. Liu, Does ChatGPT enhance student learning? A systematic review and meta-analysis of experimental studies, Comput. Educ. 227 (2025) 105224
work page 2025
-
[43]
A.R. Doshi, O.P. Hauser, Generative AI enhances individual creativity but reduces the collective diversity of novel content, Sci. Adv. 10 (28) (2024) eadn5290
work page 2024
-
[44]
V. Paananen, J. Oppenlaender, A. Visuri, Using text-to-image generation for architectural design ideation, Int. J. Archit. Comput. 22 (3) (2024) 458-474
work page 2024
-
[45]
S. Peng, E. Kalliamvakou, P. Cihon, M. Demirer, The impact of AI on developer productivity: evidence from GitHub Copilot, arXiv (2023) 2302.06590
work page internal anchor Pith review Pith/arXiv arXiv 2023
-
[46]
S. Sedkaoui, R. Benaichouba, Generative AI as a transformative force for innovation: a review of opportunities, applications and challenges, Eur. J. Innov. Manag. (2024)
work page 2024
-
[47]
E. Zhou, D. Lee, Generative artificial intelligence, human creativity, and art, PNAS Nexus 3 (3) (2024) pgae052
work page 2024
-
[48]
L. Long, X. Chen, R. Wen, T.J.J. Li, L.C. Ray, Sketchar: supporting character design and illustration prototyping using generative AI, Proc. ACM Hum.-Comput. Interact. 8 (CHI PLAY) (2024) 337
work page 2024
- [49]
- [50]
-
[51]
S.R. Cox, W.T. Ooi, Conversational interactions with NPCs in LLM-driven gaming: guidelines from a content analysis of A generative AI driven interactive narrative serious game for stress relief and its randomized controlled pilot study *Email: tingchenhsu.ac@gmail.com; 14 player feedback, in: International Workshop on Chatbot Research and Design, 2023, pp...
work page 2023
-
[52]
C. Hu, Y. Zhao, J. Liu, Game generation via large language models, in: 2024 IEEE Conference on Games (CoG), 2024, pp. 1-4
work page 2024
-
[53]
V. Kumaran, J. Rowe, B. Mott, J. Lester, Scenecraft: automating interactive narrative scene generation in digital games with large language models, in: Proceedings of the AAAI Conference on Artificial Intelligence and Interactive Digital Entertainment 19 (1) (2023) 86-96
work page 2023
-
[54]
S. Hu, T. Huang, G. Liu, R.R. Kompella, S.F. Ilhan, Y. Xu, et al., A survey on large language model-based game agents, arXiv (2024) 2404.02039
work page internal anchor Pith review arXiv 2024
-
[55]
Wenzel, Basic strategies of cognitive behavioral therapy, Psychiatr
A. Wenzel, Basic strategies of cognitive behavioral therapy, Psychiatr. Clin. 40 (4) (2017) 597-609
work page 2017
-
[56]
srcnalt, OpenAI-Unity, version 0.2.2 [computer software], GitHub, 2024
work page 2024
-
[57]
E.M. Boucher, J.S. Raiker, Engagement and retention in digital mental health interventions: a narrative review, BMC Digit. Health 2 (1) (2024) 52
work page 2024
-
[58]
I.D. Ezawa, S.D. Hollon, Cognitive restructuring and psychotherapy outcome: a meta-analytic review, Psychotherapy 60 (3) (2023) 396
work page 2023
-
[59]
K. Hammerfald, H.H. Jahren, O.A. Solbakken, The association between patient engagement and treatment outcome in guided internet-delivered CBT for anxiety and depression, Front. Psychol. 16 (2025) 1494729
work page 2025
-
[60]
J.A. Johnson, M.J. Zawadzki, F.T. Materia, A.C. White, J.M. Smyth, Efficacy and acceptability of digital stress management micro-interventions, Procedia Comput. Sci. 206 (2022) 45-55
work page 2022
-
[61]
A.Y. Bischof, T. Budig, S. Schläpfer, Y.X. Lukic, F. Schneider, P. Santhanam, et al., Long-term usage of Breeze, a gamified breathing training app, and its effect on momentary relaxation in people with cancer: cohort study, JMIR Serious Games 13 (2025) e70297
work page 2025
-
[62]
R. Taher, C.W. Hsu, C. Hampshire, C. Fialho, C. Heaysman, D. Stahl, et al., The safety of digital mental health interventions: systematic review and recommendations, JMIR Ment. Health 10 (1) (2023) e47433
work page 2023
-
[63]
G.K. Aktaş, V.E. İlgin, The effect of deep breathing exercise and 4-7-8 breathing techniques applied to patients after bariatric surgery on anxiety and quality of life, Obes. Surg. 33 (3) (2023) 920-929
work page 2023
-
[64]
C.V. Russoniello, K. O’Brien, J.M. Parks, EEG, HRV and psychological correlates while playing Bejeweled II: a randomized controlled study, Annu. Rev. Cyberther. Telemed. (2009) 189-192
work page 2009
-
[65]
Imran, Combat stress, anxiety and panic attacks: 5-4-3-2-1 coping technique, J
A. Imran, Combat stress, anxiety and panic attacks: 5-4-3-2-1 coping technique, J. Trauma Stress Disord. Treat. 9 (4) (2020)
work page 2020
discussion (0)
Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.