When scientific methods mirror stereotypes: Rethinking sex and gender differences in sport and exercise science

Girls and women are consistently less active than boys and men across the lifespan. One reason is the widespread belief that sport is a "naturally" masculine domain. This commentary argues that the scientific literature on sex differences reinforces such stereotypes through methodological biases at all stages of research. Decisions in data collection (e.g., failure to account for training status, hormonal cycles, or body composition), statistical analysis (e.g., reliance on binary significance testing), and data representation (e.g., bar plots emphasizing group means over variability) tend to accentuate between-sex differences while obscuring within-sex diversity. Such approaches mirror the mechanisms through which stereotypes operate. Yet, evidence from endurance sports and prepubertal populations shows contexts where performance differences are minimal or even reversed. We call for greater methodological vigilance, including more nuanced sampling, analyses, and visualization tools (e.g., violin, beeswarm or density plots) to better represent variability. Rethinking research practices is essential to dismantle stereotypes and promote gender equity in sport and physical activity.