Learning from refutation texts about scientific topics with analogical and causal explanations

Abstract

Previous research has suggested that refutation texts are effective in facilitating learning and revision of misconceptions and that explanations are an essential component in their efficacy. In this study, we investigated the extent to which reading refutation texts featuring an analogy as an explanatory tool rather than a causal explanation results in different encoding processes during reading, belief in misconceptions, and confidence in beliefs. Using a think-aloud methodology, we found that a higher proportion of text-based inferences was associated with analogy texts than with non-analogy texts, and a higher proportion of knowledge-based inferences was associated with non-analogy than with analogy texts. Contrary to our hypotheses, no significant differences were found in belief in misconceptions and confidence outcomes between the text conditions. We discuss the implications of these findings in light of previous research suggesting that analogies improve learning outcomes.