The Effect of Learning-Oriented Assessment on EFL Learners’ Peer and Self-Rating Accuracy of Reading Comprehension Tasks: A Mixed-Methods Study

Learning-oriented assessment (LOA) is a method that integrates assessment into the educational process, with a focus on formative feedback and collaborative engagement. However, the practical efficacy of LOA in improving the accuracy of self- and peer ratings in an English as a foreign language (EFL) context has not yet been examined. Therefore, this study, following an explanatory sequential mixed-methods design, explored the role of LOA in the self- and peer rating accuracy of intermediate EFL learners. To this end, 57 intermediate EFL learners were non-randomly divided into an experimental and a control group. Over a 10-week intervention, participants in the experimental group (n = 28) completed LOA tasks, including rubric-guided self-/peer-rating of reading comprehension and strategy-oriented feedback; the control group (n = 29) followed conventional summative assessments for the same period. Pre/posttests self- and peer rating accuracy measures (Δ scores = |learner rating – teacher rating|) were calculated using independent and paired samples t-tests. Ten semi-structured interviews were also thematically analyzed. The interviews were coded by two coders, and then the inter-coder agreement was calculated. The Cohen’s kappa coefficient of .89 indicated acceptable reliability. The quantitative analysis showed that the experimental group’s self-rating and peer rating accuracy significantly improved. The control group, however, showed no significant change. Additionally, the thematic analysis revealed five emergent themes: self-rating as a tool for metacognitive growth, peer rating as a collaborative learning opportunity, ambiguity in rubric application, emotional barriers to honest self-assessment, and desire for structured guidance. The findings indicate that LOA can significantly enhance rating accuracy.