Design and Validation of a Questionnaire on Teachers' Uses of Generative Artificial Intelligence

Synthesis: Pérez-Montesdeoca, Rodríguez-Rodríguez, Stendardi, and Fernández-Sogorb (2026) design and validate a questionnaire measuring secondary-school teachers' educational uses of generative AI. After expert judgment on an initial item pool, they administered the instrument to 486 secondary-school teachers in Spain, and confirmatory factor analysis supported a six-dimension structure — teacher management, creation of teaching materials, student assessment, student empowerment, attention to diversity, and student Motivation — with high internal consistency and adequate convergent and discriminant validity.

Key Findings

• Confirmatory factor analysis supported a six-dimension structure for teachers' GenAI use: teacher management, creation of teaching materials, student assessment, student empowerment, attention to diversity, and student motivation.
• The instrument showed high internal consistency and adequate convergent and discriminant validity (N = 486 secondary teachers in Spain).
• Teachers reported using GenAI both to optimize their professional performance and to enrich students' learning experiences.

What this means for practice

• Researchers. Measure teachers' GenAI use as a multidimensional construct with the six-dimension instrument rather than a single adoption or frequency item, so that management, materials creation, assessment, empowerment, diversity, and motivation are distinguished.
• Researchers. Re-validate the questionnaire in other countries and educational stages before comparing scores across systems, as the authors recommend.
• Researchers. Triangulate the self-report scale with institutional logs or observational data, since self-reports are exposed to social desirability and reference-frame effects.
• Teacher educators. Use the six dimensions as a needs-analysis grid that tells you which uses of GenAI (for example assessment or attention to diversity) training should target, and design and test interventions against those competency areas.

Limitations

• The sample is 486 secondary teachers in Spain recruited by non-probability convenience sampling driven by institutional access and timetabling; the authors note this can over-represent institutions open to innovation and differentially include more motivated teachers.
• All GAI use is self-reported, so social desirability and reference-frame effects mean responses may not capture actual classroom practices or the effectiveness of those uses.
• The study was cross-sectional and confined to psychometric validation; it did not examine the relationship between reported GAI use and educational outcomes such as academic performance or student engagement.
• The authors caution against generalizing the six-dimension structure to teachers in other countries or educational levels without further cross-cultural validation.

Citation

Pérez-Montesdeoca, H., Rodríguez-Rodríguez, D., Stendardi, D., & Fernández-Sogorb, A. (2026). Design and validation of a questionnaire on teachers' uses of generative artificial intelligence. Computers and Education Open, 10, 100332.