For all the research conducted about the effect of class sizes in education, only a handful of studies have focused on its impact in higher education. This despite much larger variance in class sizes in higher education (in relation to K-12), where on any given campus course enrollments may range from 1 to 1,000. So, imagine my delight when I came across research from two professors at the University of Richmond (.pdf) studying the effects of a natural experiment that took place in the business school there. For several years the dean allowed professors to “super-size” their courses from three sections of 30 students to two sections of 45 students. There was no change in the student body, in the number of students, in the courses taught, or in the faculty members. The study included data from before the super-sizing policy went in place and after it ended.
In terms of student-reported assessments of how much they learned, how they rated their instructor, and how they rated the course overall, students in classes with 30 students rated their courses higher than students in classes of 45. But the effects were minimal, and class size was not nearly as important as whether the class was early in the morning, if the student was a junior or senior, the student’s incoming GPA, and whether the student had prior interest in the course. For an experiment where the class size increased by 50%, it had surprisingly little effect on student perceptions of the course. This research is by no means definitive–if for no other reason than the outcome variable, student ratings, are inferior to actual measures of student achievement–but it’s an important step forward nonetheless.
Click Image To Enlarge
5/11/2012 | U.S. News & World Report highlights ideas from "Moving On Up: How Tuition Tax Breaks Increasingly Favor the Upper-Middle Class." 
5/11/2012 | In Diverse, Steve Burd comments on the Obama administration's plans to address student debt. 
5/11/2012 | Kevin Carey talks to Salon about the government's role in rising student debt.
{ 6 comments }
This is what we need. If each school district give every student starting with grade eight a lap top that they could only use to log into their distrect school on a broudband signell. And that distrect school is conected to all other schools in the United States. You could match students acording to there skils, so students would be equaly grounded with others in there cyber classroom. Students wouldnt be intimadated, bullied or tormited in cyber school. Who would want to drop out, i bet the drop out rate would sease. Teach students how to use computers and the internet safely and responably as a tool. A tool that can easly become an addiction in some surcomstances or get them in big trouble if there not carefull. And not let the fear of danger cause computer elitercy. There would be no need for a maultaplcation of schools or buses. Grade eight to twelve staying at home could sleep longer studies show high school students dont get the sleep they need. Foke could build schools healty and green with lots of natural light, and plants that help clean the air the students breath. Fewer bigger schools with sevrel acors of green space all around the school with lots of trees and organic hairloom vegetable guardens so the young foke can learn about there habitat and how to plant, grow, harvest, preserve, pickel, seed harvest and seed storage. You only need to buy hairloom seed the first time. Students could stay in touch with plant life and free there mind from being bombarted by gagets. Preschool students to grade seven could have more teachers to teach them. Little foke learn the best the first eleven years of there life, so i here on NPR by experts. Install a big tuch screen like the one they use on CNN in every class space, every student could have a key board at there desk that is pluged into the big screen, along with a digitel writeing pad and E book and tuch screen so they can enteract on the class space screen. Beginers could put a stinsel on the writeing pad and trace letters or numbers to help them get started with writeing. Teachers could use more of there time teaching students because everything could be recorded attomadlicly as the day gos, releaveing teachers of so much paper work. Every student would have an equal oppratunity for a great education. There is no limit to what foke can learn if they have the resorses and good drection. Every school teacher could chose any creckulum or informachion they could want.
Leonie, I’m not saying class size doesn’t matter at all, just that it is one of many factors that impact student performance. For example, prior interest in the class mattered 41 times more than class size.
The same is true in the study you cite. Class size did not meet the threshold (95% confidence) that we typically associate with statistical significance. But even if it had, things like the percentage of females and high-income students in the class and the student’s GPA mattered more than class size. (For David and John’s sake, I’ll note that the study used self-reported earnings in higher ed.)
As usual, Ms. Haimson digs a bit deeper and gets to the heart of the matter. One has to suspect bias when a writer announces how “delighted” he is about a study. In fact, this study probably doesn’t even merit comment as it is only based on student satisfaction scores at the university level and contains no actual data on student performance.
This study clearly has no application to an elementary, middle, or high school environment. The students who make it as far as college have gained the basic skills, shown success and in this environment need NO individualized attention; nor if they needed it would they be given it.
Thanks for sharing this research Chad. I would love to see what a similar study would say about learning in a chemistry class at an urban high school if its enrollment went from 30 to 45 (as many did last year).
The authors of the above cited study “find that class size has a negative and statistically significant (at the 95 percent level or greater) impact on the amount of critical and analytical thinking required in the course, the clarity of presentations, the effectiveness of teaching methods, the daily preparedness of the instructor for class, the instructor’s effectiveness in stimulating student interest, the instructor’s enthusiasm for the class, the instructor’s availability outside of class, the instructor’s respect for his/her students, the applicability of graded material to the course content, the adequacy of comments on student work, the timeliness of feedback, and even the usefulness of the text to learning. ….The only course attributes that were not statistically significantly influenced by class size were the level of course workload and the level of difficulty of the course.”
You should also check out the new study that provides robust evidence that class size matters in college in terms of both student achievement and eventual wages – and is a cost -effective reform, even for students randomly assigned to college introductory lecture classes of 100 more students. The effects are especially large for lower-income and male students:
” Our baseline results suggest that increasing class size by 20 students reduces a student’s wage by approximately 6 percent. Given this estimate, it would be hard to dismiss class size reduction as an ineffective and inefficient policy.”
CLASS SIZE AND CLASS HETEROGENEITY
Giacomo DeGiorgi, Michele Pellizzari,William Gui Woolston,
NBER Working Paper 16405, http://www.nber.org/papers/w16405
Comments on this entry are closed.