RESUMO
What we as scientists and educators assess has a tremendous impact on whom we authorize to participate in science careers. Unfortunately, in critical gateway chemistry courses, assessments commonly emphasize and reward recall of disaggregated facts or performance of (often mathematical) skills. Such an emphasis marginalizes students based on their access to pre-college math preparation and misrepresents the intellectual work of chemistry. Here, we explore whether assessing intellectual work more authentic to the practice of chemistry (i.e., mechanistic reasoning) might support more equitable achievement. Mechanistic reasoning involves explaining a phenomenon in terms of interactions between lower scale entities (e.g., atoms and molecules). We collected 352 assessment tasks administered in college-level introductory chemistry courses across two universities. What was required for success on these tasks was rote math skills (165), mechanistic reasoning (36), neither (126), or both (25). Logistic regression models predict that the intellectual work emphasized on in an assessment could impact whether 15-20% of the cohort passes or fails. Whom does assessment emphasis impact most? Predicted pass rates for those often categorized as "at-risk" could be 67 or 93%, depending on whether their success was defined by rote calculation or mechanistic reasoning. Therefore, assessment transformation could provide a path toward advancing the relevance of our courses and educational equity.