The Knewton Blog

Our monthly newsletter features edtech and product updates, with a healthy dose of fun Knerd news.

How School Holidays Affect Student Scores

Posted in Adaptive Learning on December 17, 2015 by

In The Friday Effect we looked at Knewton data showing that students do worse on Fridays. Our analysis showed that a large group of American elementary school students using an online math tool scored 4 to 5 percentage points lower on Fridays than other weekdays.

The Knewton data science team went back to look at those nearly 13,000 elementary school students to examine how the runup to a holiday affects student performance.

Because of the Friday effect, the best comparison for Thanksgiving week scores is to how the same students did on all other Mondays, Tuesdays, and Wednesdays that year. For non-holiday weeks, their average score during the beginning of the week was 74 percent. Before Thanksgiving, it was 67 percent.


That’s an even larger drop in performance than the Friday effect.

Now let’s see if there’s a similarly large drop on the Fridays before a long weekend. Might students show a similar slump before a shorter holiday?

The chart below, shows average student scores on Fridays before a two-day weekend as well as on the Fridays before Martin Luther King Jr. Day and Memorial Day.


There’s almost no difference between a normal Friday and the one before Martin Luther King, Jr., Day. However, there is a significant drop in average score before Memorial Day.

The final hypothesis we tested was whether students scores changed before Winter Break. For each weekday, the average score for the rest of the school year is in blue and the the average score for the week before Winter Break is in green.


The fall-off gets worse each day. By the Friday before Winter Break, students are scoring more than 15 percentage points lower than they would on an ordinary Friday — or about 20 points lower than on the average weekday.