Transition to Daylight Saving Time and student academic performance

The transition to Daylight Saving Time (DST) caught me out this weekend. Fortunately, I don't feel like I lost sleep as I was already exhausted from teaching this trimester. Putting aside teacher performance, the idea that losing an hour's sleep, or disruptions to the body's natural circadian rhythm, affects students' academic performance is a common working hypothesis in education circles. So, I was interested a couple of weeks ago to read this 2017 article by Stefanie Herber, Johanna Sophie Quis, and Guido Heineck (all University of Bamburg, in Germany), published in the journal Economics of Education Review (ungated earlier version here).

They used data from 22,000 European fourth-grade students from six countries, drawn from the 2011 waves of the Trends in International Mathematics and Science Study (TIMSS) and Progress in International Reading Literacy Study (PIRLS) studies (see here for more on those studies). The interesting thing about the 2011 waves is that, in the countries that Herber et al. look at, they know on what date the students were assessed, and they can identify some students who were assessed just before a change into DST, and other students who were assessed just after the change into DST. Since schools were effectively randomly assigned to when their students were assessed, this provides a useful quasi-experiment of the effect of DST on student academic performance.

Herber et al. looked at how maths, reading, and science performance differed by DST status for these students, and found that:

...students scored about 4 points lower in both math and science when tested during the week after the clock change. Given a standard deviation of about 72 and 70 points, respectively, test scores in the week after the clock change drop by roughly 6% of a standard deviation. These effects are not substantial in terms of either statistical significance or magnitude.

That quote doesn't make it particularly clear, but the effects were both tiny and statistically insignificant. In other words, DST made no difference to how well students performed on these low-stakes tests in maths, reading, or science. Herber et al. also looked at the effects on performance for eighth-graders, and again found effects that weren't consistently statistically significant and were small in magnitude (albeit on a smaller sample of countries). They conclude that:

This might be due to the fact that one hour of sleep loss is not enough to unbalance circadian clocks to the degree that performance within the following week measurably declines...

Based on our research, it is fair to say that neither parents, nor children, nor competence testing agencies (or even policy makers) have reason to worry about the alleged harmful effects of the clock change on low-stakes test performance.

So, we should probably stop worrying (if we ever were) about the effect of Daylight Saving Time on student academic performance.