The impact of teacher training on the evaluation and selection of STEM augmented reality applications and TPACK...

The digital transformation of education is reshaping the demands placed on teachers, as new competencies are required for the integration of emerging technologies such as augmented reality (AR). To utilize such digital learning resources effectively, teachers must initially be able to evaluate and select them based on their professional knowledge, which the TPACK framework conceptualizes by integrating technological, pedagogical, and content knowledge. However, it remains an open question to what extent teacher training supports the development of these competencies.

This study examines how teacher training influences prospective teachers' knowledge-based evaluation and selection of STEM-related AR applications, as well as their self-assessed digital competencies. A total of N = 305 prospective lower secondary school teachers evaluated two AR applications related to one of three STEM topics, selected the one they considered more suitable for use in the classroom and provided self-assessments of their PCK, TPK and TPACK. To explore potential differences based on training level, comparisons were made between undergraduate and graduate students. The data were analyzed using qualitative content analysis, Mann-Whitney U tests, and Pearson chi-squared tests.

Graduate students placed greater emphasis on TCK when evaluating the mathematics (p = 0.002, d = -0.58) and the physics AR applications (p < 0.001, d = -0.65), whereas undergraduates focused more on TPK in these subject assessments (mathematics: p = 0.007, d = 0.50; physics: p = 0.018, d = 0.43). Additional differences appeared within subject assessments, with the strongest effect observed for the PCK subcategory of model knowledge and use in the physics assessment (p < 0.001, d = -0.85). Under ideal conditions, undergraduates showed stronger preferences in selecting AR application in the mathematics (p = 0.030, Cramér's V = 0.195) and the biology assessment (p = 0.004, Cramér's V = 0.262), while graduates demonstrated a more balanced selection pattern. Graduates rated their PCK higher overall (p = 0.002, d = -0.37), whereas no significant group differences were observed in self-assessed TPK or TPACK.

The findings show changes in the knowledge-based evaluation and selection of AR applications, as well as in self-assessed PCK during teacher training. However, improvements in both the reference to TPACK and self-assessment of TPACK were less pronounced, indicating room for further development. This aligns with prior research suggesting that more comprehensive, model-based approaches (e.g., SQD) and stronger role modeling by teacher educators could better support teachers in effectively integrating digital resources like AR.