Skip to main content
Fig. 1 | Biology of Sex Differences

Fig. 1

From: Lower variability in female students than male students at multiple timescales supports the use of sex as a biological variable in human studies

Fig. 1

Men consistently exceed women in the variance of daily timing across semesters. Histograms with overlaid boxplots of median activity phase (men: red; women: blue; curves normalized within sex) on class days (a) and non-class days (b) find no difference on class days, but a significant delay of women relative to men on non-class days. Men are also shown to have significantly higher proportions at either extremely early or extremely late phases when compared to women (b, diagonal-indicated ranges). Histograms of standard deviation (SD) in the median activity phase across all semesters (c) reveals that low-SD individuals are significantly more likely to be women, while higher-SD individuals are more likely to be men (grey: difference of women and men). This pattern is consistent within each of the 4 semesters (d), not just across the average. The mean and standard error of the mean (SE) of activity by hour of the day across all semesters (e) reveals that men show less consolidated inactivity across individuals at night (midnight to 8am) than women, consistent with a wider range of chronotypes and individual variability observed. Conversely, women show a wider range of relative activity frequency as a population across an average non-class day (higher acrophase, lower bathyphase). Comparison of the hour-by-hour SEs of men and women’s activity (f) finds that men have higher average hourly SEs in all 4 semesters. Paired comparison of SE for the population mean of all individual’s means for each hour of the day across all four populations (24 h per average day/semester × 4 semesters = 96 comparisons) reveals that women’s SE exceeds men’s in only 4 of 96 comparisons. Shades of gray match from the bottom of f to triangles in g. *Significant; see the “Results” section for stats

Back to article page