Abstract

Feedback has become plentiful, immediate, and dialogic, and the problem of feedback has shifted to one of eliciting, comparing, evaluating, and applying it: a shift with high stakes for the critical appraisal of evidence in biomedical and natural sciences education. This conceptual article is a theoretical synthesis across three literatures of first-quartile quality: literature on feedback literacy and feedback processes for learning; scholarship on internal feedback, evaluative judgment, and self-regulation; and new research on AI-generated feedback. The article introduces the Critical Appraisal Agency Model (CAAM), which consists of a recursive four-phase model (dialogic elicitation, comparison and internal feedback, critical evaluative judgment and self-regulated enactment), two moderating layers (AI feedback literacy and hybrid regulation of learning), and one enabling condition (pedagogical and curricular design). It is formalized into six testable propositions and a series of conceptual figures, and implications for teaching, institutional policy, and research in biomedical and natural sciences education are discussed.