Unequal class sizes for all?

See all posts

Since equal class sizes don't lead to equal education for everyone, might unequal class sizes be a better option? Yes, according to a new analysis  from The Thomas B. Fordham Institute by Michael Hansen of CALDER. Overall, students may perform better if more effective teachers are assigned more students.

Using North Carolina data, Hansen finds that if more effective teachers in 8th grade classes welcomed an extra 12 students (above an average of 20 students), and less effective teachers gave up those 12 students (so they have classes with as few as eight students), it would be the equivalent of adding an extra two-and-a-half weeks to the school year - or of removing the bottom five percent of teachers. Adding only six students would still add the equivalent of almost two more weeks of school.

Importantly, both the extra students added to the effective teachers' classes and the students still in the ineffective teachers' classes would benefit - though the students already assigned to effective teachers who now have a dozen more classmates may incur a "penalty."

A big caveat: this report doesn't say much about this "penalty" for students in the supersized classes. The report describes how principals should assign classes by determining when the benefit of adding another student to the class becomes less than the cost - but says little about how to calculate that cost. After all, even effective teachers may have more classroom management problems, less time to grade work, and greater challenges building relationships with students as class sizes balloon - especially in states where the average class size starts somewhere closer to 30 students than to 20.

With the understanding that we can't 'fire our way to Finland,' Hansen presents an interesting approach to maximizing available resources.  And perhaps an approach like this might soften some resistance to performance pay, since teachers could be compensated for teaching more students, not just for being more effective.