Feedback in the Digital Classroom: How AI is Shaping Teacher and Student Feedback Processes

• In Event: Assessment Literacies in a Post-Pandemic Digital Age: An Exploration Through Six Systematic Reviews

Abstract

Teacher assessment literacy continues to expand in relation to the impact and evolution of generative artificial intelligence (AI) in K-12 classrooms. Teachers are increasing required to integrate AI into their pedagogy and assessment in ways that maintain academic integrity and assessment validity (Cope et al., 2021; Dehouche, 2021; Eaton, 2023). The aim of this paper is to specifically analyzes the existing and emerging research on how AI is shaping K-12 teachers’ formative assessment strategies, with a particular focus on feedback delivery and uptake. The research questions guiding this study are:
1. What AI applications are being used to support feedback processes in K-12 contexts?
2. How are students and teachers using AI applications to engage in feedback processes?
3. What impact is AI-driven feedback having on students’ learning and assessment experience?
4. What challenges are being reported related to AI-driven feedback practices within K-12 education contexts?

Perspective: This study takes the perspective that assessment is a socio-cognitive activity where feedback drives learning forward (Andrade & Brookhart, 2020; Holstein et al., 2023). Feedback is understood as the process of identifying where a learning is at in their learning in relation to the learning goal. Feedback can be generated by the teacher, other students (i.e., peer assessment), the student (i.e., self-assessment), or now by technology (i.e., AI-generated feedback). While supporting cognitive development, the generation, implementation, and use of feedback is situated within social contexts that either promote or negate the productive impact of feedback (Andrade & Brookhart, 2020).

Method: A systematic review process was used to identify empirical studies that focussed exclusively on ‘K-12 classroom assessment’, ‘AI’, and ‘feedback’ (or related terms). A total of 694 articles were initially identified from a search of Ebscohost, PsychInfo, and Web of Science. After initial abstract review, a total of 64 studies were selected for final inclusion in research. All studies were systematically analysed for content themes related to the research questions.

Results: While AI in summative assessment is more often discussed in the extant literature, initial results from this study are pointing to a range of applications and implementation strategies for teacher and student use of AI in feedback processes. Importantly, results are suggesting that AI is transforming traditional feedback mechanisms with greater potential for individualized and personal feedback.

Significance: The findings from this study will not only consolidate the latest research on AI and assessment, but will serve to guide teachers, policy makers, and researchers in how AI can be effectively leveraged to support newer models of feedback for student learning. The study concludes with recommendations for educators and future research.

Authors

• Christopher DeLuca, Queen's University - Kingston

• Anastasiya A. Lipnevich, City University of New York

• Therese N. Hopfenbeck, University of Melbourne

• Carmen Florentin, Graduate Center - CUNY

• Emma Carter, University of Melbourne

• Nathan H. Rickey, Queen's University - Kingston

• Joshua McGrane, University of Melbourne