New Methodologies for Examining and Supporting Student Reasoning in Physics

Learning how to reason productively is an essential goal of an undergraduate education in any STEM-related discipline. Many non-physics STEM majors are required to take introductory physics as part of their undergraduate programs. While certain physics concepts and principles may be of use to these students in their future academic careers and beyond, many will not. Rather, it is often expected that the most valuable and longlasting learning outcomes from a physics course will be a repertoire of problem-solving strategies, a familiarity with mathematizing real-world situations, and the development of a strong set of qualitative inferential reasoning skills. For more than 40 years, the physics education research community has produced many research-based instructional materials that have been shown to improve student conceptual understanding and other targeted learning outcomes (e.g., problem solving). It is often tacitly assumed that such materials also improve students’ qualitative reasoning skills, but there is no documented evidence of this, to date, in the literature. Furthermore, a growing body of research has revealed that a focus on conceptual understanding does not always result in the anticipated performance outcomes. Indeed, students may demonstrate solid conceptual understanding on one physics question but fail to demonstrate that same understanding on a closely related question. This body of research suggests that reasoning processes general to all humans (i.e., domain-general processes) may impact how students understand and reason with physics concepts. Methodologies that separate (to the degree possible) the reasoning involved in a physics problem from the conceptual understanding necessary to correctly answer that problem are necessary for gaining insight into how conceptual understanding and domain-general reasoning processes interact. In order to explore such research questions, new research tools and analysis methodologies are required. Physics education researchers pursuing these questions have begun to embrace data-collection methodologies outside of the written free-response questions and think-aloud interviews that are ubiquitous in discipline-based education research. Some of these researchers have also begun to utilize dual-process theories of reasoning (DPToR) as an analysis framework. Dual-process theories arise from findings in cognitive science, social psychology, and the psychology of reasoning. These theories tend to be mechanistic in nature; as such, they provide a framework that can be prescriptive rather than solely descriptive, thereby providing a theoretical basis for examining the interplay of domain-general and domain-specific reasoning. In the work described in this thesis, we sought to gain greater insight into the nature of student reasoning in physics and the extent to which it is impacted by the domain-general phenomena explored by cognitive science. This was accomplished by developing and implementing new methodologies to examine qualitative inferential reasoning that separate reasoning skills from understanding of a particular physics concept. In this work we present two such methodologies: reasoning chain construction tasks, in which students are provided with correct reasoning elements (i.e., true statements about the physical situation as well as correct concepts and mathematical relationships) and are asked to assemble them into an argument in order to answer a physics question; and possibility exploration tasks, which are designed to measure student ability to consider multiple possibilities when answering a physics problem. The overarching goal of these novel tasks is to explore mechanistic processes related to the generation of qualitative inferential reasoning chains and to uncover insight into the nature of student reasoning more generally. The work reported in this dissertation has yielded a variety of important results. In concert with reasoning-chain construction tasks, the dual-process framework has been leveraged to provide testable hypotheses about student reasoning and to inform the design of an instructional intervention to support student reasoning. By applying network analysis approaches to data produced by reasoning chain construction tasks with network analysis, insights were uncovered regarding the structure of student reasoning in different contexts, and the development of a coherent reasoning structure over the course of a two-semester physics course was documented. Finally, students’ tendency to explore possibilities has been, both in the literature and in this dissertation, found to impact performance on physics questions. This tendency is examined and a possible mechanism controlling this tendency has been proposed. Taken together, these investigations and findings constitute substantive advances in how student reasoning is studied and serve to open new doors for future research