Abstract:
Gaining conceptual knowledge in science is regarded as a fundamental goal in education (BMBF, 2019; KMK, 2009). However, several studies have shown only rudimentary and superficial conceptual knowledge in STEM (Science, Technology, Engineering, Mathematics) subjects (MINT Nachwuchsbarometer, 2020; Reiss et al., 2019). Conceptual knowledge is considered a crucial requirement to deeply understand new information and to transfer it to other contexts (Gruber et al., 2000; Jong & Ferguson-Hessler, 1996; Mandl et al., 1994). Generative learning strategies may trigger students to deeply process new content and to monitor their learning, which has been shown to enhance students’ conceptual knowledge and their monitoring accuracy (Brod, 2020; Fiorella & Mayer, 2016; Fukaya, 2013). In this context, one strategy received attention in both research and practice: learning by explaining to fictitious peers (Hoogerheide, Visee, et al., 2019). Generating an explanation to a fictitious peer is regarded as a beneficial learning strategy, as students engage in cognitive and metacognitive processes, which should result in higher levels of conceptual knowledge (Fiorella & Mayer, 2014; Lachner et al., 2021). However, prior research demonstrated mixed findings regarding the effectiveness of explaining to fictitious peers (e.g., Fiorella & Mayer, 2013, 2014; Fukaya, 2013; Hoogerheide et al., 2014; Hoogerheide, Renkl, et al., 2019; Hoogerheide, Visee, et al., 2019; Lachner et al., 2020). Additionally, large variances among studies were reported, which might highlight that the effectiveness of explaining to fictitious peers depends on boundary conditions (Kobayashi, 2019; Lachner et al., 2021). Such boundary conditions, however, have not yet been investigated systematically. In addition, little is known about the underlying mechanism of explaining to fictitious others. This dissertation aims to close both research gaps. In a first step, a theoretical framework model was generated based on prior research that considered implementation-related (i.e., explanatory modality, text complexity, social presence) and student-related (i.e., prior knowledge, academic self-concept) boundary conditions. Second, three underlying mechanisms of explaining were explored and included in the framework model, which were derived from previous research (i.e., retrieval practice hypothesis, generative learning hypothesis, social presence hypothesis). Lastly, the framework model was empirically tested within three studies.
In Study 1, I investigated whether text complexity moderates the effectiveness of generating oral versus written explanations regarding students’ conceptual knowledge (i.e., factual knowledge, transfer knowledge) and their monitoring accuracy. University students (N = 115) studied a complex versus a simple text and then explained the content to a fictitious peer in either oral or written form. In a control group, students engaged in a retrieval practice activity. Results revealed that explaining is only beneficial when the provided text is complex, but not when it entails a low level of complexity regarding both students’ transfer knowledge and their monitoring accuracy. Additionally, results showed that oral explaining was more beneficial than writing explanations. The explanatory effect was mediated by students’ perceived social presence as students who explained orally showed higher levels of perceived social presence, which triggered students to generate more comprehensive explanations and resulted in higher transfer knowledge. In Study 2, I conducted a replication study with an additional writing condition in which perceived social presence of the fictitious peer was induced. In contrast to Study 1, results revealed no differences among conditions regarding students’ conceptual knowledge and monitoring accuracy. In Study 3, I investigated the impact of students’ prerequisites and tested whether school students’ prior knowledge and academic self-concept moderated the effectiveness of explaining in oral versus written form. Results showed that academic self-concept, but not prior knowledge moderated the explanatory effect, as only students with low academic self-concept benefited from explaining regarding their factual knowledge. Finally, I discussed whether explaining to fictitious others is effective by conducting two additional meta-analyses. Results revealed a small positive effect of explaining on students’ factual knowledge but not on their transfer knowledge. Additionally, results of the dissertational studies showed that the explaining effect on students’ comprehension was mediated by students’ perceived social presence during explaining. However, simply inducing social presence did not result in higher learning outcomes. Finally, results revealed that oral explaining was more beneficial than writing explanations, and that students’ academic self-concept but not their prior knowledge moderated the explaining effect.
In summation, this dissertation provides a systematic investigation of potential boundary conditions of learning by explaining to fictitious peers and additionally investigated underlying mechanisms of explaining. Results of this dissertation revealed that explaining is a beneficial learning strategy to enhance students’ conceptual knowledge in STEM subjects, however, it is crucial to consider explanatory modality and students’ academic self-concept as boundary conditions of explaining.