Abstract :
[en] Introduction
This study is part of a broader effort to validate psychometric scales translated and adapted to the French-speaking Belgian teacher education context. It examines the factor structure and internal consistency of several instruments measuring preservice teachers’ attitudes toward science and educational research and their perceived scientific skills.
Method
Following the methodological recommendations on validation processes, the study focused on item development (translation, contextual adaptation, expert review), scale development (cognitive pretesting, exploratory factor analysis), and initial evaluation through confirmatory factor analysis. Six translated scales were first administered to an exploratory sample of 186 pre-service teachers, and the instruments with the most promising psychometric properties were subsequently examined on a confirmatory sample of 523 students. Exploratory factor analyses (parallel analysis, MAP, EGA, minimal residual extraction with oblique rotation), confirmatory factor analyses with WLSMV for ordinal data, and reliability indicators (McDonald’s omega) were used to evaluate the dimensional structure and internal consistency of the scales.
Results
Three scales (BISS, SESCI, and P-TATER) showed interpretable factor structures with adequate internal reliability (
ω
between 0.83 and 0.96), while three others (ATOSS, Adapted Barometer, and TPACI) did not meet the basic criteria for sampling adequacy and structural stability. The combination of exploratory and confirmatory analyses provided initial evidence for a one-dimensional solution for the BISS, a two-factor solution for the SESCI, and a three-factor solution for the P-TATER, despite strong correlations between the dimensions.
Conclusion
This study provides preliminary evidence of the psychometric quality of three instruments for assessing preservice teachers’ attitudes toward science and educational research in the French-speaking Belgian context. It also highlights the limitations and contextual fragility of the other translated scales. The results remain partly provisional, particularly for the IRS, which has only been subjected to exploratory analysis, and require further confirmatory work. Nevertheless, the study provides an initial basis for the development of a set of stable and context-appropriate measurement tools that meet contemporary psychometric standards and support research-based teacher training.
