Science Teachers' Perception of Automated Scoring Scientific Argumentation in a Classroom

This study examines science teachers' perceptions of automated scoring systems (AS) in teaching scientific reasoning in the classroom. Using a combination of diffusion theory and digital assessment literacy as an analytical framework, focus group interviews were conducted to explore teachers' attitudes, understandings, and challenges with AS. Findings suggest that teachers primarily used AS as a supplemental tool rather than a primary assessment method, relying on its insights as an additional reference throughout the course. Teachers found that understanding the concepts underlying machine learning algorithms not only increased their confidence in AS, but also inspired new approaches to integrating AI with science content. A key contribution of this study is its detailed examination of teachers' first-hand experiences with AI-enhanced learning systems, which provides insights into how AS can be effectively incorporated into science education. The findings contribute to the design of supportive AI-enhanced assessment environments that are aligned with teachers' instructional goals and pedagogical values.