Developing an attitude scale for preservice teachers towards artificial intelligence technology: A validity and reliability study

Abstract

This study aimed to develop a measurement tool to assess preservice teachers' attitudes toward artificial intelligence technology. In this study, a cross-sectional design was employed within the scope of quantitative research. The study sample was selected using a convenient sampling method and comprised 672 preservice teachers from various departments and grade levels in the Faculty of Education of a state university. The scale development process has been carried out systematically. A pool of 70 items was generated from the relevant literature and formatted as a Likert-type scale. Field experts evaluated the scale's content validity, and feedback from 15 preservice teachers was gathered to inform necessary revisions. An exploratory factor analysis of data from 315 preservice teachers revealed a three-factor structure comprising 39 items. The scale explained 53.23% of the total variance. Confirmatory factor analysis performed on data from another group of 357 preservice teachers supported the exploratory factor analysis results. The fit indices are as follows: χ²/df = 2.86; RMSEA = .072; CFI = .96; NNFI = .95; NFI = .93; IFI = .96; RFI = .92. Convergent and discriminant validity analyses revealed that the measurement model has a generally acceptable level of validity. Furthermore, the high Cronbach's alpha coefficient indicates the scale's reliability. The final scale consists of 39 items: 28 are positively worded, and 11 are negatively worded. These findings indicate that the scale is both valid and reliable, providing a robust tool for measuring preservice teachers' attitudes toward AI technology.