The Mediating Role of Generative AI Self-Regulation on Students’ Critical Thinking and Problem-Solving
Open accessThe DOI matches its Crossref record; 45 of the 45 references it cites were found in a registry or an index; 8 of the 16 sources its claims cite were checked for support.
Abstract
Within the rapid integration of AI into educational settings, understanding its impact on essential cognitive skills is crucial for developing effective teaching strategies and improving student outcomes. This study examines the influence of generative artificial intelligence (GenAI) on students’ critical thinking and problem-solving skills in higher education. Our research specifically investigates how the perceived ease of use, usefulness, and learning value of GenAI tools might influence students’ critical thinking and problem-solving skills, and whether self-regulation serves as a mediator in this relationship. Utilising a quantitative approach, we surveyed 223 students and analysed their responses using a structural equation modelling method. The results reveal that the ease of use of GenAI significantly enhances self-regulation, which in turn positively impacts both the critical thinking and problem-solving abilities of students. However, the perceived usefulness and learning value of GenAI were not found to significantly influence these skills through self-regulation. These findings suggest that, while AI tools can offer an environment conducive to developing higher-order cognitive skills, this might not necessarily translate to the enhancement of students’ skills. This research contributes to the ongoing literature on the role of technology in education by highlighting the importance of designing GenAI tools that support self-regulated learning. Furthermore, it calls for educators and developers to focus not just on the functionality of AI, but also on how these tools can be integrated into curricula to effectively support critical thinking and problem-solving. The practical implications of our research highlight the need for AI tools that are user-friendly and aligned with educational goals, enhancing their adoption and effectiveness in improving student outcomes. It is crucial for educators to integrate strategies that promote self-regulation within AI-enhanced learning environments to maximise their impact on student learning.
Overview
CoChat Summary
TL;DR
This study surveyed 223 higher-education students in China and used PLS-SEM to test whether GenAI perceptions affect critical thinking and problem-solving through self-regulation. It found that perceived ease of use significantly enhanced self-regulation, which in turn supported both outcomes, while perceived usefulness and learning value did not significantly operate through self-regulation.
Key findings
- Perceived ease of use had a significant indirect effect on critical thinking via self-regulation (β = 0.213, p = 0.000) and on problem-solving via self-regulation (β = 0.120, p = 0.001).
- Direct effects: perceived usefulness and ease of use positively predicted critical thinking (β = 0.215, p = 0.008; β = 0.205, p = 0.010), while perceived learning value did not (β = -0.001, p = 0.986). All three predictors positively predicted problem-solving (β = 0.199, p = 0.002; β = 0.237, p = 0.001; β = 0.279, p = 0.000).
- Mediation hypotheses involving perceived usefulness and perceived learning value were rejected for both critical thinking and problem-solving (H3a, H3b, H4a, H4b).
- The model explained 45.3% of variance in self-regulation, 50.1% in critical thinking, and 60.9% in problem-solving; SRMR was 0.067.
Methods
- Quantitative survey of 223 students recruited by email from 650 randomly selected undergraduate and postgraduate students at a public university in China; final sample was 67% female, 69.51% undergraduate, 29.15% master’s, and 1.35% PhD.
- Measures adapted from Davis, Alves/Zimmerman, Pintrich’s MSLQ, and Rosenbaum, using a 7-point Likert scale; items were translated into Chinese with back-translation and content-validated with 12 scholars and students.
- Data analyzed with Partial Least Squares Structural Equation Modelling (PLS-SEM) in SmartPLS, including measurement model assessment, convergent and discriminant validity, and structural model hypothesis testing.
Limitations
- Stated: the study examined only ease of use, perceived usefulness, and perceived learning value, not other AI attributes such as interactivity, feedback quality, or personalisation.
- Stated: the Chinese educational context may limit universal applicability, and future cross-cultural research is needed.
- Stated: the study primarily used quantitative methods; future research could add qualitative interviews or focus groups. Evident from the text: the design is a cross-sectional self-report survey, so causal claims and links to objectively measured skill development are not established.
Model-written summary. Check it against the paper before relying on it.
Integrity signals
Registry
Cites
Statistics
Figures


Verification
Claims the paper makes
6 Valid · 1 Not valid · 5 Not checked · 53 Own results and interpretations
The cited sources that could be checked couldn’t confirm or dispute this claim.
In the paper
“Research has demonstrated a connection between employing GenAI technologies and fostering these critical and problem-solving skills.”
§ The Mediating Role of Generative AI Self-Regulation on Students' Critical Thinking and Problem-Solving > 2.1. The Impact of AI on Skill Development
What the sources say
- Could not verifycites [12] Bailey, D., Almusharraf, N., and Hatcher, R. (2021). Finding Satisfaction: Intrinsic Motivation for Synchronous and Asynchronous Communication in the Online Language Learning Context, Springer. sourceExperimental — accuracy not yet measured
The Conclusion and Limitation and future directions sections examine motivation, online language learning, and course satisfaction; the latter mentions artificial intelligence only as a future possibility. Searches for GenAI, generative AI, critical thinking, and problem-solving skills found no passage connecting GenAI use to development of those skills. No page numbers were provided in our copy. (LLM, conf=0.77)
- Not checked — no full text of the sourcecites [3] Essien, 2024, The influence of AI text generators on critical thinking skills in UK business schools, Stud. High. Educ. source
boostedstates an association- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Could not verify
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“These three critical attributes of GenAI have been shown to significantly influence students' motivation for self-directed learning and subsequently their skill development.”
§ The Mediating Role of Generative AI Self-Regulation on Students' Critical Thinking and Problem-Solving > 2.2. The Mediating Role of Self-Regulation
What the sources say
- Partly supportedcites [28] Li, K. (2023). Determinants of college students’ actual use of AI-based systems: An extension of the technology acceptance model. Sustainability, 15. sourceExperimental — accuracy not yet measured
The Conclusions (p. 12) report that perceived usefulness and perceived ease of use directly affected three aspects of college students’ learning motivation when using AI-based systems. The paper does not establish learning value as a third attribute, specifically study GenAI or self-directed learning, or show that motivation subsequently led to skill development. (LLM, conf=0.88)
- Partly supportedcites [29] Al-Abdullatif, A.M. (2023). Modeling Students’ Perceptions of Chatbots in Learning: Integrating Technology Acceptance with the Value-Based Adoption Model. Educ. Sci., 13. sourceExperimental — accuracy not yet measured
Sections 5.1 and 5.2 report that perceived ease of use significantly predicts perceived value and perceived usefulness significantly predicts attitudes toward chatbots. Section 2.1 discusses chatbots’ potential to promote motivation and skills, but the paper does not show that the three named attributes significantly influence motivation for self-directed learning or subsequent skill development. (LLM, conf=0.84)
boostedstates a cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Partly supported
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“However, the phenomenon of GenAI producing biassed or inaccurate information, commonly referred to as GenAI hallucination, has surprisingly been observed to contribute to the enhancement of students' critical analysis skills.”
§ The Mediating Role of Generative AI Self-Regulation on Students' Critical Thinking and Problem-Solving > 2.1. The Impact of AI on Skill Development
What the sources say
- Partly supportedcites [16] Musi, E., Carmi, E., Reed, C., Yates, S., and O’Halloran, K. (2023). Developing misinformation immunity: How to reason-check fallacious news in a human–computer interaction environment. Soc. Media Soc., 9. sourceExperimental — accuracy not yet measured
The Conclusion (page unavailable) reports that users perceived improvements in critical-thinking skills after using chatbots designed to help them identify fallacies in news. It does not attribute those improvements to GenAI hallucinations or biased or inaccurate AI output, and the participants are described as users rather than students. The paper also cautions that self-reported improvements may not translate into increased ability. (LLM, conf=0.86)
plainstates a cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Partly supported
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“This observation corresponds with patterns noted in the critical thinking domain, suggesting that students' perceptions of AI's usefulness and learning value do not directly lead to improved problem-solving skills via self-regulated learning mechanisms, similar to the findings of Ghotbi et al. [39] and Chan and Hu [40].”
§ The Mediating Role of Generative AI Self-Regulation on Students' Critical Thinking and Problem-Solving > 5.2. Results of the Mediating Role of Self-Regulation
What the sources say
- Not checked — no full text of the sourcecites Ghotbi et al. [39] Ghotbi, 2022, Attitude of college students towards ethical issues of artificial intelligence in an international university in Japan, AI Soc. source
- Partly supportedcites Chan and Hu [40] Chan, 2023, Students’ voices on generative AI: Perceptions, benefits, and challenges in higher education, Int. J. Educ. Technol. High. Educ. sourceExperimental — accuracy not yet measured
Table 4 (p. 8) records students’ concern that GenAI could hinder transferable skills, including problem-solving, while the Discussion (p. 13) reports perceived learning benefits and concerns about critical thinking. The paper does not establish whether perceptions of usefulness or learning value lead to improved problem-solving skills, directly or through self-regulated learning; that pathway is not tested. (LLM, conf=0.78)
hedged- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“Our findings demonstrate that GenAI can indeed play a constructive role in enhancing students' critical thinking and problem-solving capabilities, aligning with prior research, such as that by Essien et al. [3] and Urban et al. [18].”
§ The Mediating Role of Generative AI Self-Regulation on Students' Critical Thinking and Problem-Solving > 5.1. The Impact of GenAI on Employability Development
What the sources say
- Not checked — no full text of the sourcecites Essien et al. [3] Essien, 2024, The influence of AI text generators on critical thinking skills in UK business schools, Stud. High. Educ. source
- Partly supportedcites Urban et al. [18] Urban, 2024, ChatGPT improves creative problem-solving performance in university students: An experimental study, Comput. Educ. sourceExperimental — accuracy not yet measured
The Discussion (page unavailable) reports that university students using ChatGPT produced more original, elaborated, and task-aligned solutions, supporting improved creative problem-solving performance. The paper does not establish that ChatGPT enhanced students’ critical-thinking capabilities; its Limitations and Future Directions section calls for further study of longer-term learning and metacognitive development. (LLM, conf=0.88)
boostedstates a cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
39 more claims
Create a free account to see all 65 claims and every source verdict.
- Probably not a substantial claim. 11
- The quote only partly states this claim. 1
- It may state two separate claims as one. 13
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
The source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“Metacognitive development requires explicit scaffolding rather than emerging naturally from AI interaction (Bauer et al., 2025; Yin et al., 2025; Zhou et al., 2024).”
From Reimagining writing assessment for the AI era: a systematic review on balancing AI support and authentic skill growth (2026) · cites this paper as Zhou et al., 2024
Experimental — accuracy not yet measuredThe source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“Students using AI tools show significant gains in recall, comprehension, and basic application (Abdelmagid et al., 2025; Essien et al., 2024; Zhou et al., 2024), while higher-order skills including critical thinking, analysis, evaluation, and creative synthesis prove more vulnerable to erosion through cognitive offloading (Gerlich, 2025; Zhai et al., 2024).”
From Reimagining writing assessment for the AI era: a systematic review on balancing AI support and authentic skill growth (2026) · cites this paper as Zhou et al., 2024
Experimental — accuracy not yet measuredThe source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“providing support in evaluation and inference by making it easier for learners to compare competing claims, examine trade-offs, and evidence-based argumentation tasks.”
From A Systematic Literature Review on the Pedagogical Implications and Impact of GenAI on Students' Critical Thinking (2026) · cites this paper as [15]
Experimental — accuracy not yet measuredThe source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“Zhou et al. [15] found that GenAI supports metacognitive self-regulation by encouraging students to reflect on their reasoning process or explore multiple perspectives in problem-solving”
From A Systematic Literature Review on the Pedagogical Implications and Impact of GenAI on Students' Critical Thinking (2026) · cites this paper as [15]
Experimental — accuracy not yet measuredThe source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“While GenAI can accelerate ideation, provide feedback, and simulate Socratic dialogue, excessive dependence risks undermining independent reasoning and metacognitive self-regulation [15,18,30].”
From A Systematic Literature Review on the Pedagogical Implications and Impact of GenAI on Students' Critical Thinking (2026) · cites this paper as [15]
Experimental — accuracy not yet measured
8 more claims about this paper
Create a free account to see every claim other papers make about this one.
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
Each contradiction between two passages listed here was confirmed by an independent check.
Rated minor: it would not touch a result — a value no result rests on, a p-value that stays on the same side of the significance level, or a conflict that is clerical, resolved elsewhere in the paper, or outside its own findings.
In the paper
§ The Mediating Role of Generative AI Self-Regulation on Students' Critical Thinking and Problem-Solving > 2.1. The Impact of AI on Skill Development
“Perceived usefulness is understood as the extent to which an individual perceives that using a system would reduce their effort, both physically and mentally [5].Conversely, perceived ease of use is the belief in the potential of a particular system to improve one's job performance [5].”
§ The Mediating Role of Generative AI Self-Regulation on Students' Critical Thinking and Problem-Solving > 2.1. The Impact of AI on Skill Development
“Within the context of AI in education (AIED), perceived usefulness can be seen as the likelihood of students using AI tools based on their belief in their ability to improve academic performance [6]; meanwhile, perceived ease of use is the ease associated with using AI for educational support.”
Rated minor: it would not touch a result — a value no result rests on, a p-value that stays on the same side of the significance level, or a conflict that is clerical, resolved elsewhere in the paper, or outside its own findings.
In the paper
§ The Mediating Role of Generative AI Self-Regulation on Students' Critical Thinking and Problem-Solving > 4. Results
“ranging from 0.514 to 0.755”
§ The Mediating Role of Generative AI Self-Regulation on Students' Critical Thinking and Problem-Solving > 6. Conclusions
“| Perceived usefulness | 0.543 | 0.760 | 0.514 | | | | | Problem-olving skills | 0.641 | 0.755 | 0.729 | 0.630 | | | | Self-regulation | 0.746 | 0.623 | 0.745 | 0.490 | 0.707 | |”
Rated minor: it would not touch a result — a value no result rests on, a p-value that stays on the same side of the significance level, or a conflict that is clerical, resolved elsewhere in the paper, or outside its own findings.
In the paper
§ The Mediating Role of Generative AI Self-Regulation on Students' Critical Thinking and Problem-Solving > 4. Results
“Notably, perceptions of the usefulness, learning value, and ease of use of a tool or system significantly and positively affect problem-solving skills (β = 0.199, p < 0.001; β = 0.237, p < 0.01; β = 0.279, p < 0.01, respectively)”
§ The Mediating Role of Generative AI Self-Regulation on Students' Critical Thinking and Problem-Solving > 6. Conclusions
“| H2a: Perceived usefulness → problem-solving skills | 0.199 | 0.201 | 0.064 | 3.121 | 0.002 | Support |”
Citations
Cites
Evaluating Structural Equation Models with Unobservable Variables and Measurement Error
Fornell, C., & Larcker, D. F. (1981) · Journal of Marketing Research · doi:10.2307/3151312 · Cited by 109,314
Evaluating structural equation models with unobservable variables and measurement error
Fornell, C., & Larcker, D. F. (1981) · Journal of Marketing Research · doi:10.1177/002224378101800104 · Cited by 109,190
Hu, L., & Bentler, P. M. (1999) · Structural Equation Modeling A Multidisciplinary Journal · doi:10.1080/10705519909540118 · Cited by 108,788 · cited 2×
“The Stand-ardised Root Mean Square Residual (SRMR) is 0.067, which is below the threshold of 0.08 as suggested by Hu and Bentler (1999) [38], indicating a good fi t.”
InfluentialPerceived usefulness, perceived ease of use, and user acceptance of information technology
Davis, F. D. (1989) · MIS Quarterly · doi:10.2307/249008 · Cited by 68,483 · cited 5×
“Davis (1989) [5] 2.”
InfluentialZimmerman, B. J. (2000) · Handbook of Self-Regulation · doi:10.1016/b978-012109890-2/50031-7 · Cited by 6,069 · cited 1×
“In addition, the learning value gained from the use of GenAI was assessed in this study by 4 items which we adapted from Alves [30], who originally developed them based on the work of Zimmerman [31].”
Sample size determination and power analysis using the G*Power software
Kang, H. (2021) · Journal of Educational Evaluation for Health Professions · doi:10.3352/jeehp.2021.18.17 · Cited by 2,572 · cited 1×
“The number of respondents was determined using GPower 3.1, as recommended by previous studies such as Kang [34].”
InfluentialStudents' voices on generative ai: perceptions, benefits, and challenges in higher education
Chan, C. K. Y., & Hu, W. (2023) · arXiv · doi:10.48550/arxiv.2305.00290 · Cited by 2,504 · cited 1×
“…with patterns noted in the critical thinking domain, suggesting that students’ perceptions of AI’s usefulness and learning value do not directly lead to improved problem-solving skills via self-regulated learning mechanisms, similar to the findings of Ghotbi et al. [39] and Chan and Hu [40].”
Students’ voices on generative AI: perceptions, benefits, and challenges in higher education
Chan, C. K. Y., & Hu, W. (2023) · International Journal of Educational Technology in Higher Education · doi:10.1186/s41239-023-00411-8 · Cited by 2,301 · cited 1×
“…with patterns noted in the critical thinking domain, suggesting that students’ perceptions of AI’s usefulness and learning value do not directly lead to improved problem-solving skills via self-regulated learning mechanisms, similar to the findings of Ghotbi et al. [39] and Chan and Hu [40].”
ChatGPT: Bullshit spewer or the end of traditional assessments in higher education?
(2023) · Journal of Applied Learning & Teaching · doi:10.37074/jalt.2023.6.1.9 · Cited by 1,806 · cited 1×
“While AI-generated responses and content may seem credible, they can also produce misinformation and “alternative facts” based on the prompts provided by users [41].”
A SWOT analysis of ChatGPT: Implications for educational practice and research
Farrokhnia, M., Banihashem, S. K., Noroozi, O., & Wals, A. (2024) · Innovations in Education and Teaching International · doi:10.1080/14703297.2023.2195846 · Cited by 1,384 · cited 2×
“Farrokhnia et al. [14] have demonstrated that employing GenAI tools might undermine advanced cognitive capabilities, such as critical and analytical thinking. such as critical and analytical thinking.”
Research works printed but not linked: 1
- Schunk, D.H., and Zimmerman, B.J. (2018). Handbook of Self-Regulation of Learning and Performance, Taylor & Francis.
Cited by
Izquierdo-Condoy, J. S., Arias-Intriago, M., Tello-De-la-Torre, A., Busch, F., & Ortiz-Prado, E. (2025) · Journal of Medical Internet Research · doi:10.2196/76340 · Cited by 87
“These divergent findings suggest that the effectiveness of GenAI in supporting—or impairing—critical thinking among medical students is likely influenced by several variables, including tool design, task complexity, and the degree of faculty oversight [15,23].”
InfluentialLyu, W., Zhang, S., Chung, T., Sun, Y., & Zhang, Y. (2025) · Computers and Education: Artificial Intelligence · doi:10.1016/j.caeai.2025.100383 · Cited by 68
“Enrichment [126] Using GenAI enhances student problem-solving skills.”
Xiaoyu, W., Zainuddin, Z., & Leng, C. H. (2025) · Cogent Education · doi:10.1080/2331186x.2025.2485499 · Cited by 160
Nasr, N. R., Tu, C.-H., Werner, J., Bauer, T., Yen, C.-J., & Sujo-Montes, L. (2025) · Education Sciences · doi:10.3390/educsci15091198 · Cited by 75
“This corresponds with an expanding collection of studies suggesting that AI may either impede or facilitate critical thinking based on its application (Suriano et al., 2025; Walter, 2024; Zhou et al., 2024), a notion emphasized by Robe’s reflection.”
Han, X., Peng, H., & Liu, M. (2025) · Educational Research Review · doi:10.1016/j.edurev.2025.100714 · Cited by 70
Metadata
- Authors
- Zhou, X., Teng, D., & Al-Samarraie, H.
- Affiliations
- Queen Mary University of London (GB)
- Beijing University of Chemical Technology (CN)
- School of Design, University of Leeds, University of Leeds, Leeds LS2 9JT, UK
- Corresponding author
- Teng, Da
- Venue
- Education Sciences
- Publisher
- MDPI AG
- Published
- 2024-11-27
- Volume
- 14
- Issue
- 12
- Pages
- 1302
- ISSN
- 2227-7102 (ISSN-L)
- OpenAlex
- W4404756518
- Cited by
- 119
- References
- 39
- Citations by year
- 2026: 75 · 2025: 44
- FWCI
- 44.88
- Type
- journal article
- Language
- en
- Open access
- gold
- doi.org
- Licence
- Locations
- Education Sciences (publishedVersion, cc-by)
- Queen Mary Research Online (Queen Mary University of London) (submittedVersion, cc-by)
- DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
- Funders
- Beijing Municipal Social Science Funds (23GLB016)
- Topics
- Social Sciences › Education › Artificial Intelligence in Education
- Health Sciences › Medicine › Health Informatics › Artificial Intelligence in Healthcare and Education
- Social Sciences › Education › Education and Critical Thinking Development
- Keywords
- generative AI, student self-regulation, critical thinking, problem-solving skills, perceived ease of use, perceived usefulness, value of learning, university students, AI adoption in education
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
The claim verdicts this report counts are experimental: their accuracy is not yet measured.
A language model drew the claims from the paper on 2026-09-28, under version 3 of the claim definition (prompt extract-v3).
A second model labelled each claim on 2026-09-28 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).
Cited sources are checked with a language model’s help (gpt-6-sol).
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.