Log in  |  Search

Don’t Think Class Size Affects Achievement? Think Again.

It is already obvious that class sizes are up this year — which will give School Year 2011-12 the dubious honor of being the fourth straight year of class size increases. The DOE won’t have official numbers until November, but budget cuts resulted in the loss of some 2,500 teachers this year, enrollments are rising and now we have the Day 6 class size grievance counts: nearly 7,000, up from 4,370 this time four years ago.

Will the bigger classes affect achievement? Results from just a single year suggest they will. The UFT Research Dept. looked at fourth grade, where class sizes rose an average of about one-half a child (0.47) last year. Then we divided the fourth grade into schools where class size rose more than the average, and schools where it rose less, and looked at their achievement in math. The difference was pronounced. While the majority of schools improved in math last year, schools where 4th grade class sizes rose by less than the average improved two percentage points more than schools that had larger-than-average class size increases.

School year
2010 to 2011
Growth in class size Change in math proficiency
Grade 4 (274 schools) Less than 0.47 student +5.2 points
Grade 4 (426 schools) More than 0.47 student +3.2 points

Another way to parse the same information is by dividing the group into quartiles (quarters). When we did this we found that the 25% of schools with the biggest class size increase — an average increase of 5.5 students — gained 1.87 points on the 4th grade math test from 2010 to 2011. The 25% of schools with the biggest drop in class sizes — a decrease of 4.1 students — boosted their math proficiency by almost six points (5.98 points to be exact). This suggests that the effect of raising or lowering class sizes by even more students is an even bigger change in performance. The more you reduce class size the more achievement grows.

Research has repeatedly shown that small class sizes, especially in the elementary grades, result in better performance. It allows teachers to offer more individual assistance, catching problems early and ensuring that all students are getting the lesson and mastering the concepts. Our research also suggests the opposite: when class sizes grow,  performance slides. City students will pay a price for the teacher cuts.

Notes: These results were statistically significant. For average class sizes we used a weighted average of all general education, collaborative team teaching and gifted and talented classrooms in each school, excluding special education classrooms. We tested but did not find an effect in 4th grade ELA, but chose math as a more sensitive indicator for this purpose. We would have preferred to run this analysis using more years of data, but because of the revision in proficiency cut scores in 2010, performance results cannot be readily compared with earlier years.