What Drives Achievement Gaps? The Effect of Affect and Cultural Relevance on Assessment of Numerical Skills in General Education Science Courses

Gauging student competency by performance on assessments is not as straightforward as it may seem, as performance is not a simple measure of skill. Results from the Quantitative Reasoning for College Science (QuaRCS) study reveal that affective variables such as numerical self-efficacy and math anxiety level have a strong influence on student performance, and that gaps in affect according to gender, race/ethnicity, and socioeconomic status are partially responsible for observed acheivement gaps. At the same time, the contributing affective factors and effect sizes vary greatly across demographic groups. This points to the importance of addressing multiple affective domains as part of inclusive pedagogical practices. We also find marked demographic gaps in students’ perception of the relevance of certain QuaRCS items to their lives, culture, and what’s important to them. This cultural disconnect between the assessment items and students’ lived experiences may be another contributing factor to observed achievement gaps. We will discuss our efforts to understand the effect of cultural relevancy on QuaRCS results, as well as ongoing revisions to make the assessment more cross-culturally relevant. We believe that these results have important implications for effective, inclusive pedagogy in STEM courses as well as accurate assessment of the STEM competencies of all students.