AI 전환기 대학교수자 교수역량 측정도구 개발 및 타당화

초록

This study aimed to develop and validate a diagnostic instrument for measuring teaching competencies required in higher education during the AI transition era. Preliminary items were constructed through systematic literature review, analysis of existing instruments, and expert consultation. Content validity was examined through a two-round Delphi survey with 15 experts in educational technology, educational measurement, and AI education. Construct validity was subsequently verified through exploratory factor analysis (EFA) and confirmatory factor analysis (CFA) with 312 faculty members at four-year universities, and criterion-related validity was assessed through correlation analyses with Learning Management System (LMS) log data. The final instrument comprised 48 items across five sub-factors: Core Teaching Expertise (F1), Learner-Centered Diagnosis and Support (F2), Curriculum Design and Assessment Improvement (F3), Instructional Management and Adaptability (F4), and AI-Integrated Teaching Competency (F5). EFA yielded a five-factor structure accounting for 54.22% of total variance. CFA confirmed the five-factor hierarchical model as the best-fitting final model (CFI = .93, TLI = .93, RMSEA = .056, SRMR = .053), with convergent validity (AVE = .51–.57, CR = .87–.93) and discriminant validity criteria fully satisfied. Criterion-related validity was supported through a theoretically expected convergent-discriminant pattern; notably, AI tool usage logs showed the highest convergent correlation with AI-Integrated Teaching Competency (F5; r = .61). This study contributes to the field by structuring AI-era teaching competencies into an independent construct system and validating them empirically.

키워드