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

The DOI matches its Crossref record; 32 of the 32 references it cites were found in a registry or an index; 9 of the 20 sources its claims cite were checked for support.

Enhancing service-learning through Generative AI: a mixed-methods study on educational Game design in a finance course

Open access
Wan, Kelvin, Woo, Yan Yin, Ho, G. T. S. · 2025 · Cogent Education

Abstract

Since the release of ChatGPT in late 2022, the integration of Generative Artificial Intelligence (GenAI) in higher education has posed both opportunities and challenges, especially around critical thinking and assessment integrity. This study explores how GenAI can be ethically and effectively integrated into a service-learning course on personal finance at a Hong Kong university. Students in the GenAI-supported class designed educational games for elderly participants in community centres, applying financial concepts through AI-assisted content generation and game development. Adopting a sequential explanatory mixed-methods design, we compared two student groups (GenAI-supported vs. non-GenAI) using a modified Service-Learning Outcomes Measurement Scale (S-LOMS). Quantitative results revealed significantly higher gains in Knowledge Application, Critical Thinking, and Self-Efficacy among the GenAI group. Qualitative thematic analysis of student reflections supported these findings and highlighted four key themes: enhanced knowledge transfer, increased self-efficacy, critical engagement with AI, and ethical awareness. Contrary to concerns about over-reliance, students used GenAI critically, supplementing and challenging AI outputs. These results suggest that GenAI, when scaffolded by ethical and reflective pedagogies, can enhance cognitive and metacognitive outcomes in service-learning. Implications for integrating GenAI into experiential learning and practical guidelines for ethical use are discussed.

Overview

CoChat Summary

TL;DR

This mixed-methods study integrated Generative AI into an undergraduate service-learning finance course where students designed educational games for elderly participants. Compared with a non-GenAI class, the GenAI-supported class showed significantly greater gains in Knowledge Application, Critical Thinking, and Self-Efficacy, with qualitative reflections highlighting critical and ethical engagement with AI.

Key findings

  • On post-test S-LOMS scores, the GenAI class scored higher than the non-GenAI class in Knowledge Application (M = 8.07 vs. 7.43; t(75) = 2.10, p = .04), Critical Thinking Skills (t(75) = 2.17, p = .03, d = 0.50), and Self-Efficacy (M = 7.74 vs. 6.67; mean difference = 1.07; t(75) = 2.37, p = .02, d = 0.54).
  • Paired-sample results showed larger gains in the GenAI class for Knowledge Application (mean difference = 1.25, SD = 1.78; t(34) = 4.16, p < .001, d = 0.70 vs. non-GenAI d = 0.37) and Self-Efficacy (mean difference = 1.40, SD = 2.23; t(34) = 3.72, p < .001, d = 0.63 vs. non-GenAI d = 0.04).
  • Thematic analysis of GenAI-class reflections identified four themes: enhanced knowledge transfer/knowledge application, increased self-efficacy and creative problem-solving, critical engagement with AI, and ethical awareness/responsible AI use.
  • Students reported verifying AI-generated content, noticing outdated or incorrect information, and emphasizing human oversight; the paper concludes GenAI acted as a pedagogical partner when scaffolded by ethical and reflective pedagogies.

Methods

  • Sequential explanatory mixed-methods design comparing two offerings of a 14-week undergraduate “Personal Finance and Society” service-learning course at a private Hong Kong university in Academic Year 2023/24: GenAI class (n = 36) and non-GenAI class (n = 41), with convenience sampling.
  • The GenAI class received a two-hour Game-Based Learning and GenAI workshop in Week 7 using ChatGPT 3.5 via the university’s in-house version; the non-GenAI class followed the same syllabus and project format without GenAI or the workshop.
  • Quantitative data used the S-LOMS across 11 domains and 4 categories on a 10-point Likert scale (Cronbach’s α = 0.96), analyzed with paired-sample and independent-sample t-tests in IBM SPSS 28; qualitative data were individual reflective essays from the GenAI class, analyzed using Braun and Clarke’s six-step thematic analysis with two independent coders.

Limitations

  • The paper states limitations including small sample size, single-institution scope, focus on one financial literacy course, reliance on self-reported data and reflections subject to bias, and use of ChatGPT-3.5 with known accuracy and contextual-depth limitations.
  • The paper notes that self-reported critical thinking and self-efficacy may not fully align with actual performance and calls for objective data such as peer assessments, behavioral analytics, or AI usage logs.
  • Evident from the text: the design used two intact classes with convenience sampling and no randomization, and the paper acknowledges possible context-specific factors such as instructor style or student motivation; qualitative reflections came only from the GenAI class.

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

Integrity signals

Registry

Crossref · 10.1080/2331186x.2025.2592370

Cites

32 of 32 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Verification

Claims the paper makes

7 Valid · 0 Not valid · 10 Not checked · 86 Own results and interpretations

45 more claims

Create a free account to see all 103 claims and every source verdict.

See all claims — free account
  • It may state two separate claims as one. 37
  • Probably not a substantial claim. 23
  • The quote only partly states this claim. 1

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.

4 of 4

Cited works that were retracted

Works this paper cites that were later retracted, from the retraction notices publishers deposit. Major when the notice came out before this paper was published.

1 of 1

Citations

Cites

1 retracted

Show all 31 references

Research works printed but not linked: 3

  1. Bloom, B. S., Engelhart, M. D., Furst, E. J., Hill, W. H., & Krathwohl, D. R. (1956). Taxonomy of educational objectives: The classification of educational goals. In Handbook 1: Cognitive domain (pp. 1103–1133). Longman.
  2. 10.1177/104687810223860 · doi:10.1177/104687810223860
  3. Kolb, D. A. (1984). Experiential learning: Experience as the source of learning and development. Prentice-Hall.

Metadata

Authors
Wan, K., Woo, Y. Y., & Ho, G. T. S.
Affiliations
Hang Seng University of Hong Kong (HK)
Corresponding author
Woo, Yan Yin
Venue
Cogent Education
Published
2025-11-25
Volume
12
Issue
1
ISSN
2331-186X (ISSN-L)
OpenAlex
W4416649532
Cited by
10
References
31
Citations by year
2026: 10
FWCI
8.29
Type
journal article
Language
en
Open access
Locations
  • Cogent Education (publishedVersion, cc-by)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion)
Topics
  • Social Sciences › Education › Service-Learning and Community Engagement
  • Social Sciences › Education › Artificial Intelligence in Education
  • Social Sciences › Psychology › Developmental and Educational Psychology › Educational Games and Gamification
Keywords
generative AI, service-learning, educational game design, personal finance education, ChatGPT, critical thinking, self-efficacy, knowledge application, ethical use of AI, older adults, experiential learning, higher 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 2026-10-05, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).

A second model labelled each claim on 2026-10-05 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).

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: Enhancing service-learning through Generative AI: a mixed-methods study on educational Game design in a finance course. https://app.cochat.ai/papers/10.1080/2331186x.2025.2592370
The paper
Wan, K., Woo, Y. Y., & Ho, G. T. S. (2025). Enhancing service-learning through Generative AI: a mixed-methods study on educational Game design in a finance course. Cogent Education. https://doi.org/10.1080/2331186x.2025.2592370

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.