<![CDATA[Establishing pre-intervention baseline equivalence is essential for strengthening the internal validity of quasi-experimental studies. However, baseline comparability is often underreported in educational technology research, particularly in evaluations of emerging AI-supported instructional innovations. This study addresses this methodological gap by examining baseline comparability before implementing a newly developed AI-supported gamified Project-Based Learning (PjBL) model in entrepreneurship education. The novelty of the study lies in explicitly positioning baseline assessment as a methodological foundation for strengthening the interpretation of intervention effects in AI-enhanced learning research. A quasi-experimental design was employed involving an experimental group (n = 30) and a control group (n = 26). Pre-test data were analysed using descriptive statistics and the Mann-Whitney U test because the data were not normally distributed. The experimental and control groups recorded comparable mean pre-test scores of 85.33 and 85.19, respectively. No statistically significant baseline difference was found between the groups (U = 303.00, Z = −1.467, p = .142, r = .20). These findings support the baseline comparability of the two groups and provide a stronger methodological basis for interpreting subsequent differences in learning outcomes as being associated with the intervention rather than substantial pre-existing group differences.]]>
Copyrights © 2026
