Using Different Instructional Supports to Help Students Learn Emergent Processes

de Gao, Jun

Emergence is a fundamental concept in many modern scientific theories, but emergent processes are difficult for science learners to understand. This dissertation investigated the following research questions. First, which type of instructional support is more effective in learning emergent processes while using computer simulation: receiving explanations directly (condition RE), or self-explaining (condition SE) simulation behavior. Second, can students form emergent schema without being explicitly told? Third, do students’ misconceptions about emergent processes come from a lack of the emergent schema? This study employed a 2x2 experimental design. The main independent variable is termed Cognitive Engagement, with two levels: high engagement (condition SE) versus low engagement (condition RE). The second independent variable is termed Schema, with two levels: comparing attributes of emergent and direct processes with examples (condition DES) versus only showing examples without mentioning any attributes (condition DEX). To address the first question, a pilot study was conducted among students at a U.S. graduate school of education. High-prior-knowledge participants were defined as those reporting that they had learned diffusion before. Low-prior-knowledge participants were defined as participants reporting they had never learned diffusion before. The results showed that both high-prior-knowledge and low-prior-knowledge participants who self-explained (SE) performed significantly better than those who received explanations (RE) in explaining the causal structure underlying emergent processes. To better understand which instructional support (RE versus SE) is more effective, the main dissertation study was conducted among Chinese middle school students in a classroom study. The students showed no knowledge of emergent processes before learning and were all considered as low-prior-knowledge participants. Contrary to the results of the pilot study, participants who received explanations (RE) performed significantly better than those who self-explained (SE) in understanding near transfer about diffusion and explaining the causal structure underlying emergent processes. These results might come from the differences in working memory across age, or from cultural differences surrounding the value of received instruction versus self-explanations. Regarding the second research question, middle school students who were only instructed in examples (DEX) improved significantly in understanding basic knowledge and near transfer about diffusion. Though not significantly, students in condition RE – DEX, where participants were only instructed in examples and read explanations, improved in understanding the causal structure underlying emergent processes at the posttest. These results suggested that students can form some knowledge of the emergent schema without being explicitly told. Regarding the third research question, middle school students who were instructed in the direct and emergent schema (DES) performed significantly better than those who were only instructed in examples (DEX) in understanding the basic knowledge, near transfer, and far transfer about diffusion, and explaining the causal structure underlying emergent processes. These results suggested that some misconceptions about emergent processes come from a lack of the emergent schema. However, participants in all conditions showed no difference in the categorization of the two processes. Participants more frequently mentioned emergent attributes in an example that had very similar surface characteristics with the emergent example shown in the DES condition. However, no one mentioned emergent attributes in the example that had very similar surface characteristics with the direct example shown in the DES condition. This result suggests that some misconceptions stem from difficulties participants have in correctly categorizing processes as having emergent properties or not.