Mapping the impact of generative AI in higher education: A scoping review of psychological and equity dimensions.
Open accessThe DOI matches its Crossref record; 86 of the 86 references it cites were found in a registry or an index; 16 of the 33 sources its claims cite were checked for support.
Abstract
The rapid integration of generative artificial intelligence (GenAI) into higher education has prompted growing scholarly attention toward its pedagogical and institutional implications. However, existing research has predominantly addressed technical capabilities and academic integrity concerns, while the psychological and equity dimensions remain fragmented and underexplored. This scoping review, conducted in accordance with PRISMA-ScR guidelines, systematically maps and synthesizes 87 studies published from January 2019 to March 2026 across six major databases to examine how GenAI influences learners’ cognitive and emotional processes and how its adoption intersects with educational inequalities. The findings identify five psychological themes—cognitive offloading, AI self-efficacy, emotional responses, identity threat, and trust—and six equity dimensions—digital divide, disability access, first-language inequity, socioeconomic stratification, gender and race disparities, and Global South constraints. Critically, only 7% of the reviewed studies explicitly examine the intersection of psychological and equity dimensions, revealing a significant gap in the literature. To address this fragmentation, the review introduces the Psychosocial-Equity Nexus, a novel three-layered recursive framework that integrates equity stratifiers, psychological states, and educational outcomes to conceptualize how structural inequities may be associated with psychological responses that, under certain institutional conditions, may contribute to the persistence or widening of educational disparities. The review offers actionable implications for research, institutional practice, and policy toward equitable and psychologically informed GenAI adoption in higher education.
Overview
CoChat Summary
TL;DR
This PRISMA-ScR scoping review synthesizes 87 studies published from January 2019 to March 2026 on the psychological and equity dimensions of generative AI in higher education, identifying five psychological themes and six equity dimensions. It finds that only 6 of 87 studies (7%) explicitly examine the intersection of these dimensions and proposes the tentative three-layered Psychosocial-Equity Nexus framework.
Key findings
- Five psychological themes were identified: cognitive offloading/metacognition, AI self-efficacy, emotional responses, identity threat, and trust; six equity dimensions were identified: digital divide, disability access, first-language inequity, socioeconomic stratification, gender and race disparities, and Global South constraints.
- Only 6 of 87 studies (7%) explicitly examined the intersection of psychological and equity dimensions; 54 studies (62%) were psychology-only and 27 studies (31%) were equity-only.
- The literature is geographically concentrated in high-income regions: North America (n = 34), Europe (n = 22), Asia (n = 18), Australia/New Zealand (n = 8), and Global South (n = 5); quantitatively, studies were mostly quantitative (n = 41), qualitative (n = 28), mixed methods (n = 12), and conceptual (n = 6), with tools overwhelmingly focused on ChatGPT and large language models (n = 79).
- Publication volume grew rapidly: 3 studies from January 2019 to December 2021, 5 in 2022, 18 in 2023, 31 in 2024, and 30 from January 2025 to March 2026.
Methods
- Scoping review conducted according to PRISMA-ScR guidelines, using six databases (Scopus, Web of Science Core Collection, ERIC ProQuest, PsycINFO EBSCO, ACM Digital Library, and arXiv preprints), with a search limited to January 2019–March 2026 and structured in three Boolean blocks covering GenAI, higher education, and psychological/equity terms.
- Two independent reviewers screened titles/abstracts (Cohen 0.86) and full texts (Cohen 0.91); from 2,143 records, 1,056 duplicates were removed, 1,087 records were title/abstract screened, 874 were excluded, 213 full texts were assessed, 126 were excluded, and 87 studies were included.
- Synthesis used descriptive frequency calculations plus inductive-deductive thematic analysis; no formal risk-of-bias or methodological quality assessment was performed.
Limitations
- The paper states that no formal methodological quality or risk-of-bias assessment was conducted; only English-language studies were included and grey literature/technical databases were not used, risking publication and language bias; most included studies were cross-sectional, self-report, or conceptual, so findings are associations or interpretations rather than causal links; the literature is geographically imbalanced with Global South and non-English-speaking contexts underrepresented; and only a few studies examined the psychological–equity intersection, making the Psychosocial-Equity Nexus tentative.
- Evident from the text: the exclusion criteria list repeats the same non-generative AI criterion in items (iii) and (iv), and the proposed recursive framework is based on only six integrated studies, so its feedback loops remain empirically unvalidated.
Model-written summary. Check it against the paper before relying on it.
Integrity signals
Registry
Cites
Statistics
Figures
Verification
Claims the paper makes
10 Valid · 6 Not valid · 14 Not checked · 55 Own results and interpretations
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“Blahopoulou and Ortiz-Bonnin (2025) indicate a significant difference in the attitude and benefits between users and non-users”
§ Mapping the impact of generative AI in higher education: a scoping review of psychological and equity dimensions > Results > RQ2: equity dimensions
What the sources say
- Partly supportedcites Blahopoulou and Ortiz-Bonnin (2025) sourceExperimental — accuracy not yet measured
The Results report significant differences in attitudes toward AI integration between ChatGPT users and non-users in five of six areas. The Data analysis section says perceived value and costs were assessed among users, not compared between users and non-users, so the claimed significant difference in benefits is not supported. The study concerns ChatGPT specifically rather than GenAI tools generally. (LLM, conf=0.87)
hedgedstates an association- 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.
- Rests on an assumption not shown here.
- Partly supported
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“Uppal and Hajian (2025) demonstrate that the access to AI tools is a significant determinant of productivity and academic engagement.”
§ Mapping the impact of generative AI in higher education: a scoping review of psychological and equity dimensions > Results > RQ2: equity dimensions
What the sources say
- Partly supportedcites Uppal and Hajian (2025) sourceExperimental — accuracy not yet measured
The Introduction (page unavailable) says that positive perceptions of AI tools can foster greater engagement, and the Discussion (page unavailable) links perceived usefulness and increased ChatGPT use to improved academic performance. The paper does not demonstrate that access to AI tools is a significant determinant of either productivity or academic engagement; its study examines perceptions and use rather than access. (LLM, conf=0.78)
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.
- Rests on an assumption not shown here.
- Partly supported
Every source the paper cites for this claim was checked and supports it.
In the paper
“As Liang et al. (2023) affirm, AI detection tools are more likely to punish non-native writing style”
§ Mapping the impact of generative AI in higher education: a scoping review of psychological and equity dimensions > Results > RQ2: equity dimensions
What the sources say
- Supportedcites Liang et al. (2023) source
The section “GPT detectors exhibit bias against non-native English authors” reports that detectors falsely labeled more than half of the non-native writers’ TOEFL essays as AI-generated and says their reliance on text perplexity may penalize writers with a limited range of linguistic expressions. This supports the paraphrase that detection tools are more likely to punish non-native writing style. (LLM, conf=0.94)
hedgedstates 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.
- Rests on an assumption not shown here.
- Supported
Every source the paper cites for this claim was checked and supports it.
In the paper
“Olaniyan et al. (2026) depict that epistemic inequalities are influencing the development of trust, reliance, and perceived value of AI, which is biased towards Western knowledge systems.”
§ Mapping the impact of generative AI in higher education: a scoping review of psychological and equity dimensions > Results > RQ3: intersection of psychological and equity dimensions
What the sources say
- Supportedcites Olaniyan et al. (2026) source
The Discussion (page unavailable) describes a privilege gap in reliance on generative AI and finds that its outputs privilege dominant knowledge traditions over non-Western ones. The ‘Reliability and skepticism in AI use’ section (page unavailable) links students’ trust and the value they place on AI to unequal access to other knowledge resources. (LLM, conf=0.90)
plainstates 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.
- Rests on an assumption not shown here.
- Supported
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“Although these models can be applied in identifying structural differences, they fail to sufficiently address the psychological impacts of unequal access, including low self-efficacy, increased anxiety, or identity threat, which this review has found to be important mediating processes (Pratama, 2025; Beckman et al., 2025).”
§ Mapping the impact of generative AI in higher education: a scoping review of psychological and equity dimensions > Discussion > Comparison with existing frameworks and novelty
What the sources say
- Could not verifycites Pratama, 2025 sourceExperimental — accuracy not yet measured
The Introduction discusses bias in AI-text detection by author background, not conventional digital divide models. Searches for digital divide, self-efficacy, anxiety, and identity threat found no passage evaluating those models or linking unequal access to the claimed psychological impacts; the self-efficacy mention in Materials and Methods is part of an example abstract. (LLM, conf=0.78)
- Partly supportedcites Beckman et al., 2025 sourceExperimental — accuracy not yet measured
The Discussion (p. 7) reports that students with lower self-reported digital and AI literacy were sometimes uncertain or fearful about GenAI, and that risk discourse may compound disadvantage for cautious and novice students. The paper does not conclude that conventional digital-divide models inadequately address psychological impacts, nor does it establish increased anxiety or identity threat as mediating processes. (LLM, conf=0.82)
plain- 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
46 more claims · 2 with their cited sources checked
Create a free account to see all 85 claims and every source verdict.
- It may state two separate claims as one. 27
- Probably not a substantial claim. 10
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 cited source says what the claim attributes to it.
In the citing paper
“Recent syntheses of GenAI, cognitive offloading, learner agency, and higher-order cognition likewise report mixed or conditional relations with critical thinking, autonomy, and cognitive performance (Alubthane, 2026; Ni and Li, 2026; Suazo et al., 2026; Wang et al., 2026; Zhai et al., 2024).”
From Beyond checking: verification quality, reliance calibration, and learning in generative AI-assisted higher education (2026) · cites this paper as Ni and Li, 2026
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
§ Mapping the impact of generative AI in higher education: a scoping review of psychological and equity dimensions > Methods
“This scoping review has been carried out in accordance with the standards set out by the Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension for Scoping Reviews (PRISMA-ScR) (Page et al., 2021).”
§ Mapping the impact of generative AI in higher education: a scoping review of psychological and equity dimensions > References
“66. Page M. J. McKenzie J. Bossuyt P. M. Boutron I. Hoffmann T. C. Mulrow C. D.. (2021). The PRISMA 2020 statement: an updated guideline for reporting systematic reviews.”
Citations
Cites
Scoping studies: towards a methodological framework
Arksey, H., & O'Malley, L. (2005) · International Journal of Social Research Methodology · doi:10.1080/1364557032000119616 · Cited by 36,742
The PRISMA 2020 statement: an updated guideline for reporting systematic reviews
Page, M. J., McKenzie, J. E., Bossuyt, P. M., Boutron, I., Hoffmann, T. C., Mulrow, C. D., Shamseer, L., Tetzlaff, J. M., Akl, E. A., Brennan, S. E., et al. (2021) · Systematic Reviews · doi:10.1186/s13643-021-01626-4 · Cited by 13,874
ChatGPT for good? On opportunities and challenges of large language models for education
Kasneci, E., Sessler, K., Küchemann, S., Bannert, M., Dementieva, D., Fischer, F., Gasser, U., Groh, G., Günnemann, S., Hüllermeier, E., et al. (2023) · Learning and Individual Differences · doi:10.1016/j.lindif.2023.102274 · Cited by 7,326 · cited 1×
“Simultaneously, the technostress literature provides valuable information on emotional and behavioral effects of using technologies such as anxiety, overload, and avoidance behaviors (Kasneci et al., 2023; Jovic et al., 2025; Chan, 2023).”
Artificial Intelligence in Education: A Review
Chen, L., Chen, P., & Lin, Z. (2020) · IEEE Access · doi:10.1109/access.2020.2988510 · Cited by 4,133 · cited 1×
“General surveys of artificial intelligence in education are more likely to consider technological advances and uses without addressing the psychological or equity implications in a systematic manner (Chen et al., 2020).”
Chatting and cheating: Ensuring academic integrity in the era of ChatGPT
Cotton, D. R. E., Cotton, P. A., & Shipway, J. R. (2024) · Innovations in Education and Teaching International · doi:10.1080/14703297.2023.2190148 · Cited by 3,087
GPT-3: Its Nature, Scope, Limits, and Consequences
Floridi, L., & Chiriatti, M. (2020) · Minds and Machines · doi:10.1007/s11023-020-09548-1 · Cited by 2,519
Students' 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 3×
“Similarly, the perception and interaction of the students with GenAI depends on the inequalities in the access and support systems of the context (Chan and Hu, 2023).”
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 3×
“Similarly, the perception and interaction of the students with GenAI depends on the inequalities in the access and support systems of the context (Chan and Hu, 2023).”
A comprehensive AI policy education framework for university teaching and learning
Chan, C. K. Y. (2023) · International Journal of Educational Technology in Higher Education · doi:10.1186/s41239-023-00408-3 · Cited by 1,577 · cited 2×
“The literature addressing AI ethics and governance, e.g., Meskó and Topol (2023) and Chan (2023), highlights regulatory and ethical issues but fails to explicitly consider how ethics and regulation are applied in the context of higher education learning experiences.”
Recent Advances in Natural Language Processing via Large Pre-trained Language Models: A Survey
Min, B., Ross, H., Sulem, E., Veyseh, A. P. B., Nguyen, T. H., Sainz, O., Agirre, E., Heintz, I., & Roth, D. (2024) · ACM Computing Surveys · doi:10.1145/3605943 · Cited by 1,387
Cited by
Wei, J., & Shang, Y. (2026) · Frontiers in Psychology · doi:10.3389/fpsyg.2026.1965371 · Cited by 1
“…thinking, cognitive offloading, learner agency, psychological outcomes, and epistemic autonomy, generally reporting conditional effects (Alubthane, 2026; Balart et al., 2026; Li et al., 2026; Ni and Li, 2026; 10.3389/fpsyg.2026.1965371 Suazo et al., 2026; Wang et al., 2026; Zhai et al., 2024).”
InfluentialMahyiyuddin, A. A., & Erlita (2026) · Research Square · doi:10.21203/rs.3.rs-10221208/v1
Fan, Y., Li, X., & Zhang, R. (2026) · Frontiers in Psychology · doi:10.3389/fpsyg.2026.1969217 · Cited by 4
“Only a small proportion of reviewed higher-education studies jointly address psychological and equity dimensions (Ni and Li, 2026).”
Onwuegbuzie, A. J., & Abrams, S. S. (2026) · Frontiers in Psychology · doi:10.3389/fpsyg.2026.1907909
Metadata
- Authors
- Ni, W., & Li, J.
- Affiliations
- East China Normal University (CN)
- University of Electronic Science and Technology of China (CN)
- Corresponding author
- Li, Jing
- ORCID
- Li, Jing
- Venue
- Frontiers in Psychology
- Publisher
- Frontiers Media SA
- Published
- 2026-06-08
- Volume
- 17
- Pages
- 1856854-1856854
- ISSN
- 1664-1078 (ISSN-L)
- PMC
- PMC13284694
- OpenAlex
- W7163887771
- Cited by
- 6
- References
- 84
- Citations by year
- 2026: 6
- FWCI
- 10.38
- Type
- journal article
- Language
- en
- Open access
- gold
- doi.org
- Licence
- Locations
- Frontiers in Psychology (publishedVersion, cc-by)
- PubMed (publishedVersion)
- DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
- Topics
- Social Sciences › Education › Artificial Intelligence in Education
- Physical Sciences › Computer Science › Artificial Intelligence › Ethics and Social Impacts of AI
- Health Sciences › Medicine › Health Informatics › Artificial Intelligence in Healthcare and Education
- Keywords
- generative AI, higher education, learner psychology, educational equity, cognitive offloading, AI self-efficacy, emotional responses to AI, identity threat, trust in AI, digital divide, accessibility for people with disabilities, socioeconomic stratification, racial and gender disparities, Global South, educational inequality
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 2026-10-29, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).
A second model labelled each claim on 2026-10-29 (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
How to cite
Cite CoChat for the checks on this page, and the paper for its own findings.
- This page
- CoChat. (2026). Citation check: Mapping the impact of generative AI in higher education: A scoping review of psychological and equity dimensions. https://app.cochat.ai/papers/10.3389/fpsyg.2026.1856854
- The paper
- Ni, W., & Li, J. (2026). Mapping the impact of generative AI in higher education: A scoping review of psychological and equity dimensions. Frontiers in Psychology. https://doi.org/10.3389/fpsyg.2026.1856854
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.

