Reappraising the relationships between physics students' mental models and predictions: An example of heat convection

Although prediction is claimed to be a prime function of mental models, to what extent students can run their mental models to make predictions of physical phenomena remains uncertain. The purpose of this study, therefore, was first to investigate 30 physics students' mental models of heat convection, and then to examine the relationship between their mental models and predictions of convection-related phenomena. A series of semistructured interviews was conducted to probe the participants' mental models and predictions of heat convection, and the constant comparative method was adopted for data analysis. The results reveal that the participants held a variety of mental models of heat convection, and nearly half held flawed mental models rather than a scientifically compatible one. In addition, while many participants attempted to run their mental models to make a prediction at the beginning stage of solving an interview problem, the relationship between the models and predictions became increasingly complex as the problem solving process continued. The relationships between mental models and predictions, however, could be better understood by considering the completeness of a mental model, the scale of analyzing mental models, and the retrieval of different formats of mental representations. Published by the American Physical Society under the terms of the http://creativecommons.org/licenses/by/3.0/. Creative Commons Attribution 3.0 License. Further distribution of this work must maintain attribution to the author(s) and the published article's title, journal citation, and DOI.