Class Sizes and Student Learning
A recent discussion on an ALT email circulation raised the interesting question of whether there was a threshold for class sizes, above which student learning experience diminished. Unfortunately, what followed was lots of “in my experience” Higginbotham-esque replies (with the exception of details of an interesting internal survey at NUIG), despite the original query specifically requesting evidence-based information.
A clackety-clack into Google Scholar throws up some interesting results on this topic. Unsurprisingly, the general trend is that increasing class size diminishes students’ educational experience, although the extent to which this happens seems to be luke-warm. There are two issues to consider: what is being measured to reflect something like “educational experience”; and what is the discipline.
What students think
In this regard, an interesting paper that caught my eye was one that considered the effect of class sizes in various disciplines (Cheng, 2011). This work dismisses student grades in favour of three evaluation scores derived from students: student learning, instructor recommendations, and course recommendations. Student learning was scored based on a student response to a 5-point Likert scale question “I learned a great deal from this course”. (Many of you, including myself, may be tempted to run screaming for the hills at this point. What would students know?! Cheng does make the point that she is not saying that this measure is superior to student outcomes, just a different measure. She refers to Pike’s (1996) interesting paper on student self-reporting for a discussion on this. Also, Hamermesh’s paper (2005) is worth a read for the title alone—in short, good looking lecturers get better ratings.)
Anyway, Cheng has amassed an impressive data set. “In total, the data span 24 departments, 2110 courses, 1914 instructors, and 10,357 observations from Fall 2004 to Spring 2009.” Before considering subject, on an overall level, Cheng found that for each of her three ratings, ratings fell as class sizes increased (although the smallest class sizes received both lowest and highest marks). Cheng has further used her data to generate a model to predict how student “learning” (**measured as outlined above**), instructor and course recommendations would change, so that for an increase of 50 in class size, these ratings would decrease by 1.4%, 1.3%, and 1.1% respectively. Of course, some disciplines will have smaller class sizes or may require more class-tutor interaction, so Cheng has drilled down into each discipline and determined if it is negatively or positively affected, or indeterminately effected (i.e. mixed results)
In the sciences, chemistry, biology, physics and maths were unaffected by increasing class size in this model, as were history, philosophy, and visual arts. Almost half of the disciplines surveyed were inconclusive, some showed negative effects: some engineering disciplines, political science, social science. No discipline benefits from increasing enrollment.
Cheng considers that theoretical subjects such as the sciences may have a low correlation with class size, but rather depends on other factors, such as quality of instructor or student effort. While I think there are flaws, or at best limitations to this study (as Cheng acknowledges), it does open up interesting questions. The one I am interested in is the culture of teaching chemistry, which is fiercely traditional. That this data suggests that an increasing class size would have little effect on ratings measured here in a chemistry class would in turn suggest that its teaching is still very much based on a teacher-centred philosophy. Clickers, anyone?
- Cheng, D. A. Effects of class size on alternative educational outcomes across disciplines, Economics of Education Review, 2011, 30, 980–990.
- Hamermesh, D., & Parker, A. Beauty in the classroom: Instructors’ pulchritude and putative pedagogical productivity. Economics of Education Review, 2005, 24, 369–376.
- Pike, G. R. Limitations of using students’ self-reports of academic development as proxies for traditional achievement measures, 1996, Research in Higher Education, 37, 89-114.