Student or educator? Sign up with your school email and get 6 months of Pro free.

The DOI matches its Crossref record; 24 of the 24 references it cites were found in a registry or an index; 9 of the 20 sources its claims cite were checked for support.

Multidimensional responses to generative AI in higher education: an integrative framework linking institutional governance, teaching practice, and student psychological adaptation

Open access
Kuang, Weihua, Gao, Xiaoqing · 2026 · Frontiers in Psychology

Abstract

This research establishes and validates an integrative framework to examine the multi-layered responses of higher education to generative AI, focusing on the interplay between institutional governance, pedagogical practices, and student psychological adaptation. Using a student survey sample of 477 respondents across three disciplinary clusters and linking these data to institution- and course-level governance and implementation indicators, we estimate multilevel models and identify response profiles that differentiate adaptation outcomes. The findings indicate that the congruence between governance mandates and classroom execution, bolstered by robust implementation capacity, correlates significantly with enhanced student engagement, self-efficacy, and agency, while concurrently mitigating academic anxiety. Although evidence-based integrity controls facilitate the internalization of ethical norms, they may inadvertently heighten student anxiety in the absence of explicit regulatory frameworks and accessible institutional support. Latent response profiles further reveal that a conditional-aligned, high-support configuration yields the most favorable adaptation pattern, whereas low-clarity and low-support environments correspond to reduced positive adaptation and heightened anxiety. Equity analyses indicate meaningful heterogeneity across institutional tiers and disciplinary contexts, underscoring that effective and fair GenAI governance depends on operational clarity, verification-ready assessment design, and resource-backed support.

Overview

CoChat Summary

TL;DR

This paper proposes and tests an integrative framework linking institutional GenAI governance, teaching practice, and student psychological adaptation using survey and document-coding data from 477 students across 42 institutions. It finds that governance–teaching alignment and implementation support are associated with better engagement, self-efficacy, agency, and lower anxiety, while evidence-heavy integrity controls can raise anxiety without adequate scaffolding.

Key findings

  • In multilevel models, the Governance–Teaching Alignment Index (GPAI) was the most consistent predictor: engagement β = 0.21, anxiety β = −0.18, self-efficacy β = 0.19, agency β = 0.17 (all p < .001), and ethical internalization β = 0.15 (p = .001); implementation capacity also predicted higher engagement (β = 0.16, p < .001) and lower anxiety (β = −0.12, p = .002).
  • Governance typology: “Allow Use with Conditions” had the highest GPAI mean (0.60, SD 0.09; 6 institutions, 69 students), while “Undecided/Unclear” had the lowest (0.44, SD 0.08; 14 institutions, 163 students); overall GPAI was 0.52 (SD 0.11).
  • Four latent response profiles emerged: conditional-aligned/high-support (n = 139, 29.1%) had the highest engagement (3.65 ± 0.69), self-efficacy (3.54 ± 0.64), agency (3.46 ± 0.68), and lowest anxiety (2.78 ± 0.75); low-clarity/low-support (n = 149, 31.2%) had the lowest positive adaptation and highest anxiety (3.26 ± 0.79); integrity-restrictive/high-evidence (n = 114, 23.9%) had higher ethical internalization (3.49 ± 0.58) but anxiety of 3.32 ± 0.81.
  • At course level, clear course rules were associated with lower anxiety (β = −0.20, p < .001); assessment evidence required was associated with higher ethical internalization (β = 0.13, p < .001) but also slightly higher anxiety (β = 0.06, p = .031); learning support availability was associated with higher engagement (β = 0.14, p < .001) and lower anxiety (β = −0.11, p < .001).

Methods

  • Cross-sectional student survey with stratified sampling across STEM, humanities/social sciences, and specialized disciplines; 477 valid students (170 STEM, 160 humanities/social sciences, 147 specialized) from 42 institutions, all having used GenAI for academic tasks in the current semester.
  • Student data were linked to institution- and course-level governance/implementation indicators derived through systematic text coding of official policies, teaching-center resources, FAQs, LMS guidance, and support materials; the study derived governance indicators, implementation capacity, and the GPAI.
  • Analyses included reliability/validity checks (Cronbach’s α, CR, AVE, HTMT; Cohen’s κ and Krippendorff’s α), descriptive and group comparisons, two-level multivariate models, latent class analysis (four-class solution selected), and robustness checks.

Limitations

  • Paper-stated: broad disciplinary clusters may mask within-field heterogeneity, and the sample lacks granular demographic/professional details (e.g., instructor tenure, years of experience, baseline GenAI comfort), limiting deeper interpretation and replication.
  • Evident from the text: the cross-sectional design cannot establish causal direction; psychological adaptation and teaching practice are self-reported; some governance typology cells are small (e.g., Permissibility by Default: 2 institutions, 11 students), and document coding captures formal policies/resources rather than enacted classroom practice.

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.3389/fpsyg.2026.1789506

Cites

24 of 24 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Figures

Figure 1 Study workflow and inclusion and exclusion criteria across governance documents, course materials, and student data.
Figure 1 Study workflow and inclusion and exclusion criteria across governance documents, course materials, and student data.CC BY · PubMed Central
Figure 2 Institutional governance typology of generative AI in higher education.
Figure 2 Institutional governance typology of generative AI in higher education.CC BY · PubMed Central
Figure 3 Governance to pedagogy translation map. (A) Network diagram showing alignment between institution-level governance requirements (blue nodes) and course-level pedagogical practices (green nodes). Thicker lines indicate stronger alignment. (B) Heatmap of alignment scores (0–1) across governance orientations and implementation dimensions. Higher values (yellow) represent stronger alignment.
Figure 3 Governance to pedagogy translation map. (A) Network diagram showing alignment between institution-level governance requirements (blue nodes) and course-level pedagogical practices (green nodes). Thicker lines indicate stronger alignment. (B) Heatmap of alignment scores (0–1) across governance orientations and implementation dimensions. Higher values (yellow) represent stronger alignment.CC BY · PubMed Central
Figure 4 Course-level GenAI syllabus governance templates.
Figure 4 Course-level GenAI syllabus governance templates.CC BY · PubMed Central
Figure 5 Student psychological adaptation indicators to generative AI.
Figure 5 Student psychological adaptation indicators to generative AI.CC BY · PubMed Central
Figure 6 Relationships between institutional governance/implementation indicators and student psychological adaptation outcomes. (A) Association between the institutional support capacity index and student adaptation outcome score, with points grouped by institutional tier. (B) Association between the governance implementation index and student adaptation outcome score, with points grouped by discipline cluster.
Figure 6 Relationships between institutional governance/implementation indicators and student psychological adaptation outcomes. (A) Association between the institutional support capacity index and student adaptation outcome score, with points grouped by institutional tier. (B) Association between the governance implementation index and student adaptation outcome score, with points grouped by discipline cluster.CC BY · PubMed Central

Verification

Claims the paper makes

6 Valid · 2 Not valid · 8 Not checked · 97 Own results and interpretations

57 more claims · 2 with their cited sources checked

Create a free account to see all 113 claims and every source verdict.

See all claims — free account
  • It may state two separate claims as one. 37
  • Probably not a substantial claim. 17

Create a free account to see these claims.

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.

1 of 1

Citations

Cites

Show all 21 references

Research works printed but not linked: 1

  1. Freeman, 2025, Student Generative ai Survey 2025

Metadata

Authors
Kuang, W., & Gao, X.
Affiliations
Hunan Normal University (CN)
Corresponding author
Gao, Xiaoqing
Venue
Frontiers in Psychology
Published
2026-05-05
Volume
17
Pages
1789506-1789506
ISSN
1664-1078 (ISSN-L)
OpenAlex
W7160337376
Cited by
1
References
21
Citations by year
2026: 1
FWCI
8.31
Type
journal article
Language
en
Open access
Locations
  • Frontiers in Psychology (publishedVersion, cc-by)
  • PubMed (publishedVersion)
  • PubMed Central (publishedVersion, cc-by)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
Topics
  • Social Sciences › Education › Artificial Intelligence in Education
  • Health Sciences › Medicine › Health Informatics › Artificial Intelligence in Healthcare and Education
  • Social Sciences › Education › Digital Transformation of Higher Education
Keywords
generative AI in higher education, institutional governance, teaching practice, student engagement, self-efficacy, student agency, academic anxiety, institutional support, response profiles, disciplinary differences, assessment design

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 2026-10-10, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).

A second model labelled each claim on 2026-10-10 (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: Multidimensional responses to generative AI in higher education: an integrative framework linking institutional governance, teaching practice, and student psychological adaptation. https://app.cochat.ai/papers/10.3389/fpsyg.2026.1789506
The paper
Kuang, W., & Gao, X. (2026). Multidimensional responses to generative AI in higher education: an integrative framework linking institutional governance, teaching practice, and student psychological adaptation. Frontiers in Psychology. https://doi.org/10.3389/fpsyg.2026.1789506

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.