Educational Prompt Engineering Self-Efficacy Scale (Ed-PESS): Scale Development and Psychometric Validation

Pre-service teachers' confidence in designing pedagogically purposeful prompts for generative AI tools remains an underexamined construct in teacher education. This study developed and validated the Educational Prompt Engineering Self-Efficacy Scale for Pre-service Teachers (Ed-PESS) to address the absence of a psychometrically sound instrument measuring this domain. Scale development followed Boateng et al.'s three-phase framework, grounded in a systematic literature review and expert review procedures. Two independent samples of pre-service teachers were recruited: an exploratory factor analysis (EFA) sample (n = 300) and a confirmatory factor analysis (CFA) sample (n = 305). EFA using principal axis factoring with Promax rotation produced a four-factor structure explaining 64.60% of total variance. CFA with Yuan-Bentler robust correction confirmed acceptable model fit. The final 26-item scale comprises four dimensions: Basic Prompt Engineering, Pedagogical Prompt Engineering, Ethical Prompting Practice, and Continuous Professional Development. Internal consistency was high across all subscales. The Ed-PESS provides a valid, reliable instrument for assessing prompt engineering self-efficacy in pre-service teacher education. It supports formative curriculum evaluation and targeted intervention design.