University students' diurnal learning-directed behavior is strongly influenced by school start times with...

School start times impose constraints on sleep-wake timing that may influence academic achievement. We used large university archived datasets to test the hypothesis that larger differences in timing of students' diurnal learning behavior on school days relative to non-school days would be associated with lower grades.

Diurnal learning-directed behavior was examined in 33 645 university students by analyzing their learning management system (LMS) login rhythm. We tested the associations between the phase-difference in students' behavioral rhythm on school days versus non-school days with grade point average, LMS-login phase on non-school days (LMS-login chronotype), and school start time. We also tested the chronotype-dependent effects of school start times on diurnal behavior to determine whether students obtained better course grades when their first class of the day was in synch with their LMS-login chronotype.

Students whose LMS-login rhythm was more than 2 hours earlier on school days had significantly lower grades than their peers. The change in LMS-login phase was larger in students with a later LMS-login chronotype and for earlier school start times. Minimal changes in LMS-login phase and higher course grades were observed when students' first class of the day was aligned with their LMS-login chronotype.

Our findings demonstrate that school start times have a profound impact on students' diurnal learning behavior with implications for grades. Universities can potentially improve learning by starting school later to minimize differences in diurnal learning behavior between school days and non-school days.