School start times and academic achievement - a systematic review on grades and test scores

Abstract

School start times have been at the centre of many scientific and political debates given the accumulating evidence that bell times are generally too early, and thus lead to an epidemic of sleep restriction in the student population. Recent media attention has conveyed the message that later school starts not only improve sleep but also result in better academic achievement. Several studies have been recently published on this topic requiring a comprehensive review of the results to clarify the relationship between later school start times and academic achievement to inform the general public and policy makers. To this end, we conducted a systematic review of the current literature on school starting times and academic achievement in middle and high school students, considering grades and standardised test scores as achievement measures. We followed the PRISMA guidelines for searching, including, and reporting relevant literature and identified 21 studies for detailed analysis. Evidence quality of included studies was assessed with a pre-defined risk of bias assessment using modified items from the GRADE scheme and ROBINS-I tool. About half of the reviewed studies reported no (positive or negative) effect of delaying school times on grades and test scores, while the other half reported either mixed or positive results. Given the strong heterogeneity of included studies, we grouped them according to various characteristics, such as academic outcomes, dose of delay, evidence quality, or study design to identify potential hidden effects. Despite this, we could not identify any generalisable effect beyond single studies as to whether delaying school times has clear beneficial effects on academic performance. Given that grades and scores determine future career trajectories and predict future success, the question whether school start times contribute to academic achievement is of great interest for the general public and needs to be further clarified. Mechanistically, it is very likely that improved sleep leads to or mediates improved cognitive performance and learning, but definitive conclusions on whether this also translates into better grades and scores across all students requires better evidence at this stage. Importantly, this does not preclude other positive outcomes of later start times such as improved sleep (quality), motivation or learning but draws attention on current gaps and shortcomings. To this end, we also highlight critical methodological aspects and provide suggestions to increase the evidence-level and to guide the direction of research in future studies.

Are flexible school start times associated with higher academic grades? A 4-year longitudinal study

“…Thus, we cannot preclude that we missed a subtle effect in our sample but any such effect is likely extremely small. This is also in line with several other studies that were unable to find any effect or meaningful improvements [28].…”

“…We have not been able to identify obvious explanations for this finding in exploratory analyses, except for the fact that weekend sleep was much more variant and backed by fewer data points than schoolday sleep, pointing towards a potential chance finding. A likely explanation for our null-finding for the other sleep parameters is a possible lack of power in our sample of 157 students (even though we have >16,000 longitudinal grades) given the small effect sizes previously identified (ranging around <0.1 SD; see [28]). A second possibility is that the time frame students were exposed to the new system was too short (exposure length) or that the delay was too little or infrequent (dose) in our study.…”

“…However, the evidence is not conclusive whether delayed school start times can ameliorate this pressing health and performance problem. Many previous studies suffer from study design limitations, outcome variables are not comparable, and long-term studies that track individuals over time are rare [28]. Here, we studied whether a novel timetabling system -a daily chosen flexible school start -has the potential to improve academic grades via improved sleep.…”

“…However, we argue that such simple pre-post analysis of aggregated grades is not suited to answer this complex question -although this has frequently been done using cross-sectional data. Studies on grades that performed proficient analyses, such as mixed regression models [57], quantile regression models [58] or difference-in-difference approaches [59][60][61] accounting for available confounders provided mixed results and mostly small effect sizes (for a systematic review see [28]). Nonetheless, positive effects of delayed school start times on academic achievement have been widely proclaimed -bound to raise falsely high expectations in parents and teachers.…”

“…Thus, several studies previously tried to evaluate the effects of delayed school start times on grades or scores. These studies yielded very mixed results, probably due to differences in study designs, interventions, exposure times and outcome measures [28]. However, all interventions assessed were static changes in school start times whereas the possibility to make school start times flexible has been largely overlooked.…”