Exploring Gender Differences in Persistence: Analyzing Test-Taking Behaviors in High-Stakes Assessments

Effort and persistence are critical determinants of students' academic success, skill acquisition, and future life outcomes. As such, these non-cognitive factors play a key role in discussions of social mobility. Research on gender differences in effort and persistence has shown that female students tend to be more self-disciplined and exhibit traits like conscientiousness. Studies on cognitive endurance in low-stakes tests, such as PISA, have also found that female students generally outperform males in persistence, regardless of the subject. However, research on student effort remains limited due to challenges in methodology and data availability, particularly in measuring effort accurately. Additionally, most existing studies on effort rely on low-stakes tests, which may not accurately reflect how students perform when real incentives are present. This study addresses those limitations by using a novel measure of effort from economics—test persistence. In tests where question order is randomized, the difficulty remains constant on average, allowing performance to be separated into cognitive and non-cognitive components. Initial performance reflects cognitive skills, while a decline in performance throughout the test signals persistence.

Focusing on gender differences, this study uses data from online exams conducted during the 2019/2020 academic year at Universidad Carlos III de Madrid. By leveraging the platform's ability to randomize questions, the research provides a bias-free measure of persistence. The objective is to explore gender differences in both test performance and persistence in high-stakes university exams, where real incentives to sustain effort exist, thus addressing the gaps left by previous studies that relied on low-stakes contexts.