Not only do larger classes -- up to a max of about 25, in most cases
-- work better than small ones, because of the benefits of student/teacher
interaction -- score-diverse classes work better than homogeneous
This is a surprise to a lot of people who think classes should
be "grouped by ability," but decades of irrefutable research
demonstrates that in heterogeneous classes, everyone learns better.
Obviously, lower scoring students can learn from the classroom discussions
with higher-scoring students. What's not so obvious is that
higher-scoring students will learn from the discourse with lower-scoring
students as well. A lower scoring student might ask a question that
a higher-scoring one would never think to ask -- but will benefit
from hearing the answer.
Our student surveys every year demonstrate no correlation whatsoever
between class size and score increase. What's interesting,
though, is that our surveys do demonstrate that every year, our
average score increases go up -- AND SO DID OUR CLASS SIZES for
the first six years of our operation. Almost looks as if larger
class sizes lead to greater score increases!