The Causal Effects of More Time

Often, when we cite research that purports to identify a connection between time and learning, we suggest that the findings are correlational, not causal. In other words, the research may have uncovered a trend where students perform at higher levels in schools or classrooms that have more time, but, because there are many other factors not considered in the analysis (like instructional quality or curriculum materials), we can only surmise that this one factor of time is associated with higher academic performance, but not the cause of it.

Well, now there is a just-released study that is just about as close to providing a causal connection between time and learning as one might hope to get in the real world.  (Apologies for the wonky detail of what follows.) The analysis examined a large set of schools in Israel that had, as a result of a revised funding formula, increased instructional time, but had not undertaken any other major educational or organizational changes.  With this sample (of over 900 schools), the author was then able to compare test scores in English, science and math in the 2002 – 03 school year to scores in the 2004 – 05 year to measure the effect of more time on outcomes.  Undertaking this analysis, the author presented a number of really important findings, but I wanted to highlight two in particular:
Finding #1: The more time schools had increased instructional time in a particular subject, the better students performed.  Here is a chart detailing the findings.  (The estimate is in standard deviation units.)
Finding #2: When students are disaggregated by gender and by socioeconomic status (level of parental education is used as a proxy here), the effect of increasing instructional time on student outcomes varies considerably from group to group, depending on subject.
Here’s what I think these two findings together really mean, and this research perfectly reflects NCTL’s own view of the role of time in learning. On the one hand, we know definitively from this data that, in general, more time will lead to more learning, but, on the other hand, time operates differently, depending upon how it is used—in this case, what subjects are taught—and who the learner is.  Time matters, but not absolutely.
And one final point worth mentioning.  As the author of this study notes, the findings on time can be used as a “benchmark” to test how well other interventions (e.g., pay for performance for teachers or professional development) measure up to the intervention of more instructional time.  Given how relatively large the impact of more time has been on the performance of these tens of thousands of Israeli students, these yet-to-be assessed interventions certainly have a high bar to surpass.