AI-Assisted Problem-Based Learning to Enhance Computational Thinking and Digital Problem-Solving Skills among Vocational Higher Education Students

An Empirical Study

Abstract

The rapid development of artificial intelligence (AI) has transformed the ways students access information, solve problems, and construct knowledge in technology-mediated learning environments. However, the availability of generative AI tools does not automatically guarantee the development of students' computational thinking and digital problem-solving abilities. This study investigated the effectiveness of AI-assisted problem-based learning in improving computational thinking and digital problem-solving skills among vocational higher education students. A quasi-experimental pre-test and post-test control group design was employed. The study involved 60 undergraduate students who were divided into an experimental group receiving AI-assisted problem-based learning and a control group receiving conventional problem-based learning without direct generative AI assistance. The intervention was conducted over eight instructional sessions. Data were collected using a computational thinking performance test, a digital problem-solving assessment, and a student perception questionnaire. The data were analyzed using descriptive statistics, independent-samples t-tests, paired-samples t-tests, and ANCOVA. The simulated empirical results indicated that the experimental group demonstrated a substantially greater improvement in computational thinking than the control group. The experimental group increased from a pre-test mean of 55.13 to a post-test mean of 84.27, whereas the control group increased from 54.60 to 72.13. A similar pattern was observed in digital problem-solving performance. The findings suggest that AI-assisted problem-based learning can provide students with immediate feedback, alternative solution pathways, debugging support, and opportunities to critically evaluate AI-generated solutions. Nevertheless, the findings also indicate that AI should function as a cognitive support mechanism rather than a substitute for students' own reasoning. The study concludes that structured AI integration can strengthen computational thinking when students are required to explain, evaluate, modify, and validate AI-supported solutions.