Examining the Effect of Self-explanations in Distributed Self-assessment

Abstract

Self-assessment is a twofold activity consisting of self-evaluation and self-explanation, which are considered imperative metacognitive strategies in learning science. Although the self-explanation effect has been scaffolded in numerous learning systems, it remains unclear whether the effect can still occur to students in a voluntary setting of learning such as remote learning that requires self-regulation to persist in making progress. Furthermore, it is inconclusive what students’ behavioral patterns can be when they exercise self-evaluation and self-explanation overtime. In this study, we investigate the effectiveness of self-assessment and the embedded self-explanation by dissecting semantic elements in the explanations in a multilevel analysis. The result showed that the low-performing students were challenged by the complexity of topics, which resulted in an increased error rate when they ventured into more learning opportunities. However, the self-explanation effect might occur to them and improved their performances, especially when they reflected on the content of questions that were relevant to the concepts. In summary, this study provides an insight into effective self-assessment. Specifically, it shows that students can potentially improve performances by writing compact and relevant explanations over time.