Meta-analyses of Peer Assessment and Affective Outcomes: Motivation, Self-efficacy, and Anxiety

Having classmates evaluate other classmate's work and give ratings, comments, or both (peer assessment, PA) can improve the student's learning, possibly via motivation, self-efficacy, and anxiety. This meta-analysis of 66 studies of 8,550 participants explored PA's effects on motivation, self-efficacy, and anxiety and tested for moderators. The random effects model results showed that PA raises motivation (g = 0.393; 24 effect sizes [ESs]) and self-efficacy (g = 0.616; 30 ESs) while reducing anxiety (g = -0.608; 27 ESs). The PET-PEESE results showed that the corresponding bias-adjusted mean effects were 0.171, 0.417, and -0.202, respectively. PA effects on motivation were positive for studies with experimental and control groups but negative and smaller for studies with only pretests and posttests for single groups. PA effects on self-efficacy were larger in later grades (high school or college). Anxiety showed no moderation effects. Assessment type in the control condition, academic subject, task, assessment feedback, anonymity, rating criteria, rater training, number of assessors per assignment, and assessment mode (paper vs. computer) did not moderate these effects. Hence, educators should capitalize on these PA benefits and adjust for these specific situations.