Generative and active engagement in learning neuroscience: A comparison of self-derivation and rephrase

Abstract

It is crucial to identify cognitive mechanisms that support knowledge growth. One such mechanism that is known to improve learning outcomes is generative processing: the construction of novel information beyond what is directly taught. In this study of college students, we investigate the learning outcomes associated with the generative process of self-derivation through integration, the integration of multiple related facts to generate novel information. We compare the effects of self-derivation versus an active rephrase control condition on retrieval, application, and organization of neuroscience classroom content. In the self-derivation condition, learners were prompted to generate inferences based on integration of two explicitly-taught facts. In the rephrase condition, learners were explicitly provided these inferences and asked to rephrase them. We found few overall differences between learning manipulation conditions. However, we found that, regardless of the learning manipulation condition to which learners were exposed, learners generated their own information on some trials. This generation predicted success on retrieval and application of learned information. Further, self-derivation, when successful, led to particularly high rates of retrieval when compared with active rephrase. These findings inform theory on generative processing, and demonstrate that self-derivation is a mechanism of knowledge growth that may be useful for retrieval.