Investigating Instructional Strategies To Support and Elicit Organic Chemistry Students' Reasoning

Success in organic chemistry requires developing reasoning skills and learning relevant concepts. Achieving these learning outcomes poses challenges for students; furthermore, supporting students with learning organic chemistry is often challenging for instructors. Because of the challenges associated with teaching and learning organic chemistry, students often engage in rote learning strategies, such as memorization, which may preclude meaningful learning. Hence, to better support students’ success in organic chemistry, research is necessary to explore how novel instructional strategies can promote students’ meaningful learning. Furthermore, it is important to investigate how pedagogical approaches can elicit students’ reasoning—rather than simply eliciting the outcome or final solution of students’ problem-solving—so instructors can better understand their students’ thinking. The research presented herein describes studies on instructional approaches intended to support students’ learning and elicit students’ reasoning. Most of the included research focuses on writing-to-learn, a pedagogy that uses writing assignments to support students’ meaningful learning and conceptual engagement. This dissertation also provides insight into using machine learning to analyze students’ responses to organic chemistry writing-to-learn assignments for the purpose of providing automated, formative feedback. This research was guided by a variety of research questions which seek to provide insights that can transform organic chemistry instruction at the undergraduate level. The dissertation opens with a review which synthesizes the existing research on students’ mechanistic reasoning and how students describe and explain organic reaction mechanisms. The following two chapters address the research question of how specific instructional strategies—a mobile device application and case-comparison problems—can elicit students’ reasoning about organic reaction mechanisms. The next four chapters focus on writing- to-learn, broadly addressing research questions regarding how these assignments can promote students’ meaningful learning and engagement while eliciting students’ reasoning about organic chemistry concepts and reaction mechanisms. The final two chapters specifically explore the question of whether machine learning models and automated feedback tools can be developed to analyze student writing and provide formative feedback for students’ responses to organic chemistry writing-to-learn assignments. Several theoretical and analytical frameworks grounded the presented research, including cognitive perspectives on learning and writing, theories of engagement and meaningful learning, and mechanistic reasoning frameworks drawn from the philosophy of science literature. Interviews, surveys, and responses to writing assignments were data sources, which were analyzed using qualitative and quantitative methods. Findings indicate how different instructional approaches can elicit different features of students’ reasoning when solving problems in organic chemistry. These studies suggest that even when the outcome of students’ problem-solving is the same, students often use varying approaches that may or may not be aligned with appropriate chemical thinking. Furthermore, each study reveals specific challenges students may face with different concepts and reaction mechanisms, offering implications for instructors to better support students’ learning. Findings also indicate the value of writing-to-learn to promote meaningful learning in organic chemistry by supporting both cognitive and affective engagement. Results from the studies using machine learning demonstrate the feasibility of using automated analysis of students’ writing to provide students with automated, formative feedback. Altogether, the studies on organic chemistry writing-to-learn assignments provide evidence of the value of this pedagogy for engaging students’ learning with challenging concepts while also providing a means to elicit rich data that reflects students’ reasoning. This research provides implications for instructors seeking to better elicit the depth of students’ reasoning while promoting conceptual learning.PhDChemistryUniversity of Michigan, Horace H. Rackham School of Graduate Studieshttp://deepblue.lib.umich.edu/bitstream/2027.42/175694/1/fieldmw_1.pd