Assessing artificial intelligence competence in vocational education: validation of the AICo scale and certification-based differences
TL;DR
The certification-group findings should be interpreted as exploratory rather than as evidence of substantively meaningful or causal effects, and future research should combine self-report and performance-based assessment, employ representative samples, and directly examine the individual and institutional factors associated with vocational teachers' AI competence.
Abstract
This study aimed to validate the six-dimensional artificial intelligence competence (AICo) instrument among vocational teachers in Indonesia and explore differences in perceived AI competence across teacher certification groups. Using a quantitative cross-sectional design, data were collected through convenience sampling from 622 Indonesian vocational teachers, comprising uncertified teachers (UT), certified teachers (CT), and teachers enrolled in the Professional Teacher Education programme (PT). Confirmatory factor analysis (CFA) was conducted to evaluate the higher-order measurement model, Cronbach's alpha and composite reliability were used to assess internal consistency, the Kruskal–Wallis test was employed to compare observed AICo scores across the three groups, and hierarchical regression examined certification-group differences after limited adjustment for age. The findings yielded three main results. First, the CFA supported the satisfactory fit, reliability, convergent validity, and discriminant validity of the six-dimensional AICo model. Second, participants reported a relatively high overall level of perceived AI competence (M = 3.78). Third, UT and PT reported slightly higher observed AICo scores than CT. However, certification status explained only 0.8% of the variance in perceived AICo, indicating very limited explanatory value, while age was not significantly associated with the outcome. Thus, the moderate rank-based effect (e2 = 0.075) reflects a consistent distributional shift rather than meaningful individual-level prediction. The certification-group findings should therefore be interpreted as exploratory rather than as evidence of substantively meaningful or causal effects. Future research should combine self-report and performance-based assessment, employ representative samples, and directly examine the individual and institutional factors associated with vocational teachers' AI competence.
