Tabriz University of Medical Sciences Research and Development in Medical Education 2322-2719 15 1 2026 01 01 Reliability of Peer and Self-assessment Across learning domains in Competency-Based Medical Education for Formative Assessment: An Observational Study 33367 33367 10.34172/rdme.33367 EN Sangeetha Kandasamy https://orcid.org/0000-0001-8458-512X Thilakavathi Kothandan Shivkumar Gopalakrishnan Yogambal Muthureddy Vijay Anto James Ganesh Kumar Puthisigamani Journal Article 10.34172/rdme.33367 2025 07 22 2025 10 06 Introduction: In Competency-Based Medical Education, formative assessments are integrated throughout the curriculum to offer continuous feedback, promoting self-regulatory learning and helping learners to identify the areas for improvement. Peer and self-assessment play a crucial role in CBME by fostering collaboration, self-reflection, critical thinking, and professionalism. Assess the reliability of peer and self-assessment across different learning domains in a CBME framework for formative assessment. Methods: This observational study, conducted from January 2024 to December 2024 at a Government Tertiary Care Hospital, involved a universal sampling method with 100 second-year medical students. Assessments addressed knowledge through assignments, psychomotor skills through vitals and systemic exams, and affective skills through communication and interpersonal abilities. Evaluations by peers, self, and faculty were conducted using an internally validated tool, with anonymity maintained throughout. Results: Participants were aged 18–21 with a male-to-female ratio of 1:1.2. ANOVA tests showed no significant difference in knowledge scores between self, peer, and faculty assessments (ICC=0.708; P=0.106). Significant differences were found in psychomotor (P=0.001; ICC=0.483) and attitude (P=0.001; ICC=0.234) domains, with students and peers overestimating or underestimating performance. Bland–Altman plots confirmed good agreement for knowledge, moderate for psychomotor, and poor for attitude between self, peer, and faculty ratings. Conclusion: The study found moderate reliability in the knowledge domain but poor to no reliability in the attitude and psychomotor domains, with self and peer assessments tending to overestimate performance. These findings underscore the need for more training and structured assessment strategies to enhance the accuracy and consistency of self- and peer-assessment in medical education.