Angelina Hadriani
Sistem Komputer S1 (Kampus Kota Serang), Universitas Pamulang

Published : 1 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 1 Documents
Search

Causal Machine Learning to Estimate the Impact of Using Generative AI on Students' GPA And Skill Retention Angelina Hadriani; Ade Sumaedi; Layli Ana; Sugiyanti Sugiyanti; Yuni Yulistiani
Al'adzkiya International of Computer Science and Information Technology (AIoCSIT) Journal Vol 7, No 2 (2026)
Publisher : Al'Adzkiya

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.55311/aiocsit.v7i2.411

Abstract

The rapid adoption of Generative Artificial Intelligence (GenAI) in higher education raises questions about its effects on academic performance and durable skills. This study evaluates the adjusted relationship between GenAI usage, GPA Change, and Skill Retention Score using 50,000 student records. Weekly GenAI Hours is treated as the exposure, while academic, behavioral, institutional, and psychological characteristics are included as adjustment covariates. A cross-fitted Double Machine Learning workflow is supported by overlap diagnostics, continuous-treatment sensitivity analysis, quartile comparison, and exploratory heterogeneous-effect analysis. High GenAI usage is associated with higher GPA Change (ATE = 0.0328; 95% CI: 0.0297–0.0358) and Skill Retention Score (ATE = 0.8098; 95% CI: 0.5733–1.0464). The quartile comparison shows the same direction. However, each additional weekly GenAI hour is associated with lower skill retention (ATE = −0.1096; 95% CI: −0.1298 to −0.0895). Thus, greater usage intensity does not necessarily indicate deeper learning. Because the data are observational, the estimates remain conditional on measured covariates. Universities should promote GenAI as a structured learning aid while preserving independent reasoning and sustained skill development.