Snow Days May Not Affect Student Performance

Winter has caused schools across the country to cancel school numerous times for snow days. Multiple closures force school districts to reevaluate their schedules to make up for the lost academic time.  A recent study from a Harvard University researcher found that those makeup days are not always necessary.

The study was done by Joshua Goodman, an assistant professor of public policy at Harvard’s Kennedy School of Government. He found:

That school closures due to inclement weather do not actually affect student achievement as measured by scores on the standardized Massachusetts Comprehensive Assessment System. Individual absences do.

The Massachusetts Department of education asked him to “crunch numbers” to decide whether or not snow days affect student achievement. Earlier research on the importance of instructional time provided mixed results.

The results from previous studies varied widely. Some found that the loss of learning over summer was larger for low income children (Cooper et al., 1996), but another study found that summer learning doesn’t even show up in all data (Fryer and Leavitt, 2009).

Another study found that the most successful schools have longer school days (Fryer and Dobbie, 2009; Abdulkadiroglu et al., 2009; Hoxby et al., 2009), yet another study completed by the Massachusetts Expanded Learning Time initiative found there was little impact on test scores when schools added 300 extra hours a year (Abt, 2010).

Goodman noticed one thing seemed to consistently be missing, student data.

Education researchers have underutilized one important set of variables contained in most state administrative data sets, namely attendance data.

So he decided to look into the data for students in Massachusetts from 3rd to 10th grade between 2003 and 2010. He looked into the school closures, individual absences, and standardized test scores.  He said that he “started thinking more broadly about instructional time, and disruptions to that time.”

Valerie Strauss of the Washington Post reported on his findings. Goodman’s research found that either the number of closures does not affect instructional time, or that schools are well prepared for dealing with snow days.

Schools do not do well when only a few students are absent. Estimates from the study suggest that absences account for “8-20% of the achievement gap between poor and non poor students. “

The results support a model that suggest the biggest challenge of teaching is managing students and that schools may not be prepared for those types of disruptions.

Of course, schools have little control over the weather. Other sources of variation in absences may be more interesting from a policy perspective: illness, transportation problems, domestic trouble. Constructing instruments from these is, however, tough.

02 14, 2014