<b>Generative Artificial Intelligence (GenAI) Meets Assessment: Experimental Insights into Teacher Candidates’ Attitudes and Acceptance </b>

Source abstract

This experimental study examined the impact of pre-service teachers' development of performance tasks utilizing differentiated generative artificial intelligence tools—textual, visual, and video-based—on their attitudes towards generative artificial intelligence (AI) and their acceptance levels of AI. A quasi-experimental approach was utilized to address the research issues, incorporating one control group and two randomly allocated experimental groups. Ninety-four pre-service teachers voluntarily participated in the study. The groups were categorized based on three distinct generative artificial intelligence tools utilized by the students in the development of a performance task. As a results of analyses, a significant increase in both positive attitude and acceptance level was reported in the Control group using text-based GenAI; an increase only in positive attitude in the Experiment 1 group; and a decrease in negative attitude in the Experiment 2 group. Engagement with differentiated generative AI tools resulted in notable alterations in in-group attitudes and levels of acceptability. This study underscores the essential requirement for pre-service teachers to implement a nuanced and diversified strategy for AI integration, acknowledging the advantages and limitations of different AI tools and their capacity to affect pre-service teachers' attitudes and acceptance in various manners.