Evaluating the impact of an AI-integrated learning platform on student performance: a quasi-experimental study

M.A. Yermaganbetova, Arailym Ashimbekova, Dinara Kaibassova, Riza Akhitova, Elmira Dyussembina

2026

Full text Open access: Gold Metadata (CSL-JSON)
DOI 10.3389/feduc.2026.1792353
Data confidence 70 medium
How this score is calculated

It is built from checkable properties of the record itself: whether the article has a DOI, how many independent sources confirmed its data, whether the required fields are filled in, and how recent the latest confirmation is. Disagreement between sources and a missing identifier lower it. A value of 80 or above means the record was published without further checking; 60 to 79 means it is published at reduced confidence, with a manual check advisable. A record verified by an editor by hand is never pushed below 90 by a recalculation. This scores our data about the article, not the article itself.

Abstract

The integration of artificial intelligence (AI) into higher education has intensified interest in adaptive learning environments that support personalized instruction and self-regulated learning. However, empirical evidence regarding the pedagogical effectiveness of fully integrated AI-based platforms in practice-oriented, skill-intensive university courses remains limited. This study examines the effect of an AI-integrated learning platform on academic performance, cognitive engagement, and self-regulated learning, employing a quasi-experimental design. In this study, cognitive engagement is conceptualized as a key behavioral and cognitive manifestation of self-regulated learning within an AI-supported instructional environment. 120 undergraduate students enrolled in a university-level 3D modeling course participated in the study, forming an experimental group ( n = 62) and a control group ( n = 58). The intervention lasted 8 weeks. Data were collected through pre- and post-tests, rubric-based practical assignments, final course grades, and a mixed-format satisfaction questionnaire. Quantitative analyses included paired t -tests, Bowker’s test of symmetry, effect size calculations, and confidence intervals. Qualitative data were examined using thematic analysis. The results reveal statistically significant improvements in post-test performance and practical task proficiency in the experimental group ( p < 0.05), accompanied by higher levels of perceived autonomy, adaptive support, and self-regulated learning. The findings suggest that a fully integrated AI learning ecosystem can function as a metacognitive scaffold, enhancing strategic learning behaviors and cognitive engagement in applied digital skills education. The study contributes empirical evidence to ongoing discussions on the pedagogical role of AI in higher education and outlines implications for instructional design and future research.