"Effects of Generative AI-Assisted Learning on Critical Thinking Skills among Higher Secondary School Students".
This study examined the effects of Generative AI (GenAI)-assisted learning on the critical thinking skills of higher secondary school students. A quasi-experimental pre-test and post-test design with a non-equivalent control group was used. The sample consisted of 59 Grade XII students of Garrison Academy Kharian Cantt (Army Public School and College), Punjab, Pakistan: 28 girls in the experimental group and 31 boys in the control group. For one month, the experimental group was taught with GenAI tools (Claude and DeepSeek) and the control group was taught with traditional methods. Critical thinking was measured before and after the teaching period with a 30-item Critical Thinking Questionnaire based on Facione's (1990) six skills: interpretation, analysis, evaluation, inference, explanation, and self-regulation. The data were analysed with paired-samples t-tests, independent-samples t-tests on gain scores, and analysis of covariance. The total critical thinking score of the experimental group rose from M = 72.71 (SD = 5.11) to M = 121.25 (SD = 7.79), t(27) = 29.02, p < .001, d = 7.37, and that of the control group rose from M = 68.68 (SD = 7.75) to M = 77.58 (SD = 8.03), t(30) = 8.15, p < .001, d = 1.13. The mean gain of the experimental group (48.54 points) was significantly larger than that of the control group (8.90 points), t(57) = 20.21, p < .001, d = 5.27, and the analysis of covariance, which adjusted for the pre-test score, gave the same result, F(1, 56) = 466.24, p < .001, partial η² = .893. All six skills improved more in the experimental group. For classroom practice, the result suggests that guided GenAI activities, in which students compare AI answers, check them against the textbook, and give their own reasons, are worth adding to regular teaching under the teacher's supervision. The effect sizes were unusually large, the experimental group also started higher than the control group (p = .023), and the pre-test reliability of the questionnaire was low (α = .29). The findings should be read with care because the groups were intact classes that also differed in gender. Keywords: Generative AI, Critical Thinking Skills, Higher Secondary Students, Quasi-Experimental Study
Authors
- Nazish Andleeb Babar Hussain
Publication Details
- Journal
- Perspectives in Social Sciences
- Published
- 2026-11-17
- DOI
- https://doi.org/10.5281/zenodo.24311706
- Primary Topic
- Artificial Intelligence in Education
- Type
- article
- Field-Weighted Citation Impact
- 0.00