Citizens in all roles need to be able to construct causal explanations of scientific phenomena and engage in coherent argumentation, supported by evidence and sound reasoning. These skills are needed in many contexts, from making everyday decisions to addressing complex global issues. One goal of instruction in this area is to get past assessment questions that can be answered using only memorized rules (e.g., S N 1 reactions are favored with tertiary halide substrates), toward answers that require reasoning with evidence ( why is an S N 1 pathway favored under certain conditions?). Previous work has suggested that these reasoning skills may not be effectively taught, practiced, or learned in organic chemistry courses or directly assessed on exams. In this study, we explored students’ ability to construct scientific arguments using evidence in the context of organic reaction mechanisms, specifically when comparing two proposed reaction mechanisms, making a claim as to which was most plausible, and providing reasoning. We used a qualitative coding scheme to analyze (1) the features and concepts students discussed in their arguments on exams, including the concept of granularity, (2) the connections (links) made between features and concepts, (3) the modes of reasoning revealed in the arguments, and (4) how the ideas and claims were explicitly compared. We found that ∼60% of arguments with correct claims established causal relationships between the relative stability of the carbocation intermediates and the relative activation energy barriers to the formation of these intermediates. However, the vast majority of students did not go to sufficient granularity to meet the expectations in the course, which they could have done by invoking Hammond’s postulate and hyperconjugation in their arguments. Incorrect claims were often supported by causal arguments but these arguments were incorrectly based on drawing a causal relationship between the relative steric hindrance of the starting materials and activation energy. Over 60% of students provided a linear causal mechanistic argument to justify their claim, independent of whether their claims were correct. Over 90% of arguments at least partially compared key concepts explicitly, with over 20% doing so completely. The findings suggest that the majority of students in this study understood the need to provide cause-and-effect relationships to justify their answer, beyond identifying reaction features and their effects; however, students may have struggled to identify which features of those molecules were relevant .
Stored only when both papers are in the corpus — 109 citations exist upstream.