Developing and using scientific models is an important scientific practice for science students. Undergraduate chemistry curricula are often centered on established disciplinary models, and assessments typically provide students with opportunities to use these models to predict and explain chemical phenomena. However, traditional curricula generally provide few opportunities for students to consider the epistemic nature of models and the process of modeling. To gain a sense of how introductory chemistry students understand model changeability, model multiplicity, the evaluation of models, and the process of modeling, we use a construct‐mapping approach to characterize the sophistication of students' epistemic knowledge of models and modeling. We present a set of four related construct maps that we developed based on the work of other scholars and empirically validated in an undergraduate introductory chemistry setting. We use the construct maps to identify themes in students' responses to an open‐ended survey instrument, the models in chemistry survey, and discuss the implications for teaching.

Explanations of phenomena in chemistry are grounded in discussions of particulate-level behavior, but there are limitations to focusing on single particles, or as an extension, viewing a group of particles as displaying uniform behavior. More sophisticated models of physical processes evoke considerations related to the dynamic nature of bulk solutions, in which an ensemble of molecules exists with a distribution of values that vary with respect to different parameters (e.g., speed, kinetic energy, etc.). Viewing phenomena as a varied population instead of a homogenous solution has been identified as a foundational idea that is critical for reasoning in chemistry, but little work has investigated how students reason about these ideas and how instructors can support students in viewing phenomena as a distribution of states. In this qualitative study, during semi-structured interivews twelve undergraduate general chemistry students were provided with frequency distribution graphs (number of molecules vs. speed, number of molecules vs. kinetic energy) and were asked to provide explanations and make predictions. The design and analysis of this study was informed by coordination class theory, a model within the knowledge-in-pieces perspective of cognition that defines a concept as a combination of approaches for obtaining information (read-out strategies) and a cluster of knowledge elements used to draw conclusions (causal net). Framing the varied population schema as a coordination class, this work focuses on the interaction between features students attended to in distribution graphs and the ideas they discussed. Analysis indicates students have productive resources for reasoning about a varied population in general terms, but these ideas are not necessarily activated when interpreting graphs, as reflected in the students’ readout strategies. Moreover, we posit that one of the barriers toward interpreting distribution graphs was the inappropriate application of covariational reasoning. As a practical consideration, we encourage interested instructors to review the Appendix, which provides a short summary of the main findings and suggestions for practitioners.

Chemical kinetics is an important topic that is reinforced across the undergraduate chemistry curriculum, but previous research indicates students tend to have difficulty developing a sophisticated understanding of reaction rate. In this qualitative case study, we characterized how two students conceptualized reaction rate in the context of reaction coordinate diagrams. Analysis involved using the knowledge-in-pieces perspective to model reaction rate as a coordination class. In short, a coordination class is a type of concept that involves the combination of an inferential net (a group of knowledge elements used to draw conclusions) and extractions (observations made related to the target concept). Between the two students we noted multiple distinct operationalized definitions for reaction rate; using the language of coordination class theory, these can be described as concept projections. Moreover, results indicate metonymy served as a cognitive construct that guided students' reasoning, in which a word or phrase was used to reference a larger knowledge system. On the basis of the analysis, we discuss the importance of metonymy in supporting students' problem solving related to reaction rate. As a practical consideration, we encourage interested instructors to review the Supporting Information for this work, which provides a short summary of the main findings and suggestions for practitioners.

To engage meaningfully with scientific models, undergraduate students must come to understand what counts as a scientific model and why. To gain a sense of the characteristics that undergraduate chemistry students ascribe to scientific models, we analyzed survey data that address students’ ideas about both model criteria in general and criteria related to specific models of chemical phenomena. The findings suggest that undergraduate general chemistry students possess some productive and some intuitive ideas about the characteristics of scientific models but may not have systematic or coherent conceptions about models across contexts.