EUVIMEDEUVIMED entdecken
Mixed-Methods-StudieOpen Access · CC BY-NC 4.0

Trust, verification, and reworking in generative-AI-supported MBA comprehensive examination preparation: A daily diary mixed-methods study

Qinjie Shen, Zhang Zhao, Wynn Arunrugstichai

Computers in Human Behavior Reports · 2026-12 · 24 · 101368

DOI: 10.1016/j.chbr.2026.101368 ↗

Kurz gefasst

Mixed-Methods-Studie in der Zeitschrift Computers in Human Behavior Reports, 2026-12. Verfasst von Qinjie Shen und 2 weiteren Autor:innen. Die Arbeit nennt 41 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY-NC 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Generative AI is increasingly embedded in higher education, but less is known about how AI-supported studying changes across a concentrated high-stakes preparation period. This explanatory sequential mixed-methods study followed 110 International MBA students for 21 consecutive days before a comprehensive examination. Primary analyses used Bayesian multilevel intensive longitudinal models with linear study-day adjustment. Perceived preparation progress increased systematically across the 21 days, whereas exam pressure declined. After this temporal trajectory was modeled, an apparent positive autoregressive carryover in progress was no longer credible; pressure showed a small negative residual autoregressive association, and same-day AI use was not clearly associated with either progress or pressure. On AI-use days, state trust, verification, judgment-based reworking, and progress all increased over study day. However, average trust-to-regulation, regulation-to-progress, and indirect trust-to-progress pathways were not statistically credible after time adjustment. A complete-case sensitivity analysis produced the same inferential pattern. Random-slope estimates for the verification-to-progress and reworking-to-progress relations were imprecise, precluding strong claims of between-student heterogeneity. Interviews nevertheless showed that the same diary labels captured substantively different practices: externally anchored versus AI-to-AI checking, and conceptual reconstruction versus surface editing. Taken together, the findings show that common preparation-period trends must be separated from within-person pathways and shift attention from whether students verify or rework more to how those practices are enacted.

Zitieren

APA

Shen, Q., Zhao, Z., & Arunrugstichai, W. (2026). Trust, verification, and reworking in generative-AI-supported MBA comprehensive examination preparation: A daily diary mixed-methods study. Computers in Human Behavior Reports, 24, 101368. https://doi.org/10.1016/j.chbr.2026.101368

Vancouver

Shen Q, Zhao Z, Arunrugstichai W. Trust, verification, and reworking in generative-AI-supported MBA comprehensive examination preparation: A daily diary mixed-methods study. Computers in Human Behavior Reports. 2026;24:101368. doi:10.1016/j.chbr.2026.101368

BibTeX
@article{Shen2026,
  author = {Shen, Qinjie and Zhao, Zhang and Arunrugstichai, Wynn},
  title = {Trust, verification, and reworking in generative-AI-supported MBA comprehensive examination preparation: A daily diary mixed-methods study},
  journal = {Computers in Human Behavior Reports},
  year = {2026},
  volume = {24},
  pages = {101368},
  doi = {10.1016/j.chbr.2026.101368}
}
Kostenlos starten →