Class-Size: What Works Best?

Our readers and guests talked about the impact of class size on student learning.

Feb. 21, 2007

Class-Size: What Works Best?

Guests: Douglas N. Harris, assistant professor of educational policy, University of Wisconsin-Madison; and Nancy Flanagan, a 30-year teaching veteran who is now a full-time doctoral student in educational policy at Michigan State University.

Kevin Bushweller (Moderator):
Welcome to today’s chat about the impact of class size on student learning.

What does the research say about the impact of class size on student achievement? What works best? And what educational factors are more important than class size?

Our guests will address these and other questions during this chat.

Question from Ray Kircher, Parent:
With all the effort to reduce class size to benefit students, what about the students who benefit from a quality teacher in a larger class size? This large class education for students dedicated to education is the reason many students who excel are unable to learn in a college environment. Small classrooms are more damaging to quality students because the small class environment is not realistic. That is what quality students want, reality.

Nancy Flanagan:
I think your assumption--that there are quality students who thrive in larger classes--needs some unpacking. One key issue in class size is (to borrow a cliche’) that one size doesn’t fit all. Some subjects and developmental levels may well be more efficiently taught in large groups. First-grade reading instruction, on the other hand, would be seriously and negatively impacted by larger class size--even for the most dedicated and attentive kids.

What about the distinct possibility that focused and accomplished teaching, in the very early years, will turn many kids from homes where learning is not a priority into what you are calling quality students?

I know some college students prefer large lectures, where information is absorbed, then tested, and feel that this model is “real learning"--compared to smaller seminars involving discussion, evaluation, argument. Smaller class sizes at all levels that put more onus on the individual learner do make people who prefer anonymity uncomfortable. Constructing knowledge in teams is a 21st century skill, however--something worth considering.

Question from Lexa Kandola, Admin. Specialist, Bay Farm Montessori Academy:
While class size is of the utmost importance, student/teacher ratio also plays a large part (e.g., 30 students is more acceptable when there are 2+ teachers as is done in a Montessori environment).

Douglas N. Harris:
The results from Project STAR suggest that having aides in the classroom is unimportant in terms of test scores, though almost every teacher I’ve ever spoken disagrees.

It’s also important to consider whether it would be better to use the money that would go into hiring an aide and use the money to hire more teachers and reduce class size. If an aide is paid half of what a regular teacher is paid, then a school could use this approach to reduce class size by 25 percent or more. I’m not saying this is absolutely the right thing to do (especially with a Montessori model), but these are the types of trade-offs we have to bear in mind.

Question from Debbie Miser, Lee U, Edu student:
Hello, I am currently a 50-year-old student working toward a degree in elementary ed. While observing in the 6th grade math dept, I have noticed how much more the teacher is able to accomplish with a smaller class size versus the size of the class during my elementary years. Example: 35-40 per class during the 70s versus 20-25 by todays standards. What do you preceive as the answer to overcrowed class rooms and where do we need to make changes as a society to enforce good learning habits no matter what the class size is? Thank you and good luck with your doctoral studies.

Nancy Flanagan:
Good luck to you, too, Debbie--and welcome to the world’s most rewarding profession.

I don’t think there’s a single answer to overcrowded classrooms. Class size is as much an economic issue as an instructional issue, so we will always be trying to balance the expense of lowering class size with other resource allocation issues. There will always be tradeoffs. Is it worth the expense of lowering class sizes one or two pupils per classroom in a district, for example, if the consequence is no new technology? Optimum class size is sometimes a judgment call.

I think the more critical part of your question is your observation that society needs to take a greater interest in education, and encourage students to value the gift of free public schooling. You’re talking about changing minds and hearts (not to mention media and politics)--but promoting a better-educated nation is a tremendous goal, especially for a novice teacher. All the best to you.

Question from Steven Forman, Teacher, Ramapo High School, East Ramapo Central School District:
Does class size really matter? I’ve had classes of 10, 18, 24, 28, 36 (NYC) and 40 (NYC) and I managed them all the same way with the same results - High percentage of passing state exams. Isn’t it all in the control?

Douglas N. Harris:
Class size does seem to have a modest effect on the average, but this doesn’t necessarily mean that it matters in all cases. For example, you might be espcially good at handling disciplinary issues or providing differentiated instruction even in a large class setting. Other teachers may only be effective in a small class setting and this may be why we see the modest effect on the average.

Also, it sounds like you’ve been teaching students in different schools. It may be that your students in the small classes started off at lower levels than the students in the large class setting (NYC), meaning that your small class students actually made greater gains than you think. One of the challenges of identifying the class size effect is separating the role of class size from the characteristics of students who tend to end up in small classes. It’s possible that, if you accounted for this by looking at the test score gains, then you’d see some effects.

Question from Paul M. Marino, Ph.D. Professor of Education, Delaware Valley College:
Class size isn’t as big an issue as the quality of teacher in that classroom. The better teacher will make 35 work. Is there any research tying these two issues?

Nancy Flanagan:
There’s plenty of research indicating that effective teaching is much more critical than class size. In fact, the Tennessee STAR study, considered one of the most trustworthy pieces of research on the benefits of lower class size, found that good teaching was more than twice as important as class size when examining measurable student learning.

I am hard pressed to think of a research design, however, that will accurately measure other important things that emerge when considering class size. How do we put a number on the value of personal attention, more chances to get feedback on your writing--or 3 kids in your reading group instead of 6? How do we measure the effect of custom-tailored, differentiated learning made possible by smaller numbers of students? Or even something as simple as more time for parent conferences? While it’s true that a top-notch teacher will adapt to almost any conditions, shouldn’t we consider reasonable class sizes an investment in children?

Question from Jonita Stepp-Greany, reitred Florida State University:
I am curious to know whether any of the research that purports to show that class size does not make that much difference have looked at the qualitative nature of the enrollment of the class. In my experience, only a handful of problem students in the class can have an inordinate effect on both the quantity and quality of learning in the class. Nevertheless, the larger the class, the larger the potential for the inclusion of more of these kinds of students. Another learning factor involves the kind of learning that is going on. For example, we know that project-based or problem-based learning seems to have longer lasting effects and the potential for learning inter-related skills than learning facts or skills in isolation from context; however, it is my experience that teachers avoid conducting these kinds of learning experiences when the class is larger, due to the fact that their is less control during these situations and hence more potential for serious disruptions.

Nancy Flanagan:
I think you’ve just hit on a key issue here: a small class is not necessarily an “easier” class. In my own experience, I have had very large classes (65+) that were thoroughly engaged in making music and small classes that were a constant challenge. Teachers do adjust instruction by class size to “make it work.” How do we use these ideas to think about new, more flexible conceptions of instructional groupings, rather than rigid class size language?

Question from Barbara Lovejoy, Executive Director of Generación Floreciente in SLC, UT:
As an educator and Hispanic youth advocate I realize the value of smaller class sizes, especially for English Language Learners. When we already have a shortage of teachers (especially qualified teachers to teach ELLs), I feel there might be more cost-effective measures we could take rather than mandating something that would require more teachers and space. Cross-age tutoring, including Coca-Cola Valued Youth Program, has proven to be extremely effective in many ways for both the tutors and the tutees. Also, trained volunteers are extremely valuable. I am anxious to know your thoughts and insights about this. Thanks!

Douglas N. Harris:
A common theme to my answers is that we need to pay attention to cost-effectiveness. To put this in perspective, I’ve shown in one of my studies that if schools had the same pupil-teacher ratios as they did in 1970, then we could use the money to give every teacher in the country a 40 percent raise. We’re talking about a lot of money. That doesn’t mean we shouldn’t do it, but it does mean that we need to carefully consider the alternatives.

One reason I give the teacher salary example is that you mentioned that you’re already facing a shortage of quality teachers, as are many high-minority schools. Raising salaries could help address that.

I don’t know much about the specific programs you mentioned, but progams such as cross-age tutoring are inexpensive and if they prove successful, certainly make sense, especially in low-resource schools.

Gracias para tu pregunta.

Comment from Mary Healey 35 year teacher veteran:
More importantly than class size is the philosophy of education that we adopt in this country. Size should depend on the type of student as well as the nature of the curriculum being studied. Also, the age of the student,is itelementary,middle secondary or college? All of the other teacher qualitites are important,but before this answer can be given there are many variables that must be addressed. If we actually believe that children learn by “doing” then it is obvious that the numbers must be lower to accommodate this"doing-ness” activity. If we do not believe in this philosophy, then with all the high tech at our disposal,numbers do not very much matter.

Question from Hayes Mizell, Distinguished Senior Fellow, National Staff Development Council:
For teachers who have little hope of gaining significantly smaller classes, is trial-and-error the only way they can learn to manage their classes to enhance student leaning? How can more effective professional development help teachers and students succeed in spite of classes of 25 to 35?

Nancy Flanagan:
Hayes, you almost answered your own question here. Most teachers have little input into class size or flexible grouping of students to improve instruction (an issue in its own right)--but I hate to see thoughtful building of a teaching practice referred to as “trial and error”. All worthy practitioners learn from experience--but that experience is greatly enhanced by ongoing professional learning which gives teachers an intellectual framework to analyze their own practice.

A consistent theme in the class size literature is that effective teaching trumps class size. When professional development provides positive intervention in practice, teachers do a better job of handling larger groups of kids. The key is job-embedded, just-in-time professional development, which is hardly the norm. I’m also wondering if you haven’t posed a false dichotomy--shouldn’t we be aiming, always, for optimum efficiency at reasonable cost?

Question from Bob Frangione, Educator/Parent:
Asian schools, as in Japan and China, have class sizes generally well over 30 students to a room, yet class management does not appear to be a problem. What is the key to their success, if any, in working with large numbers?

Douglas N. Harris:
It’s important to be careful when making international comparisons. The “best” education system in one country is probably not the best for another. In this case, part of the issue is that students in many high scoring countries have cultural norms that lead students to defer to the authority of teachers, producing classrooms with fewer disruptions. So, in those countries, small classes aren’t as necessary for creating an orderly learning environment.

Question from Leonie Haimson, Exec. Director, Class Size Matters:
There is often a assumed tradeoff between teacher quality and class size, and yet none of the studies in California or elsewhere have shown that class size reduction actually led to a loss in teacher effectiveness. The latest study to reveal this is:" Identifying Effective Teachers” by Gordon, Kane and Staiger, April 2006 at

What the class size studies do show is that smaller classes lead to to lower teacher attrition rates, as was found in CA, NY and elsewhere. Shouldn’t then class size reduction result in a more experienced and effective teaching force over time?

Nancy Flanagan:
I’m not sure that lower teacher attrition rates will necessarily lead to more effective teaching over time. More experienced, perhaps, as teachers are more satisfied with their work and remain in teaching. It may also be difficult, with aggregated data, to tease out whether individual (sometimes underprepared) teachers who went to work when California suddenly needed a massive influx of teachers changed the overall level of effectiveness. Did students with effective teachers, in the newly small classes, score better and mask the effect of new teachers? There need to be some interventions--professional development, collaboration--to improve instruction if we are aiming for more uniform effectiveness in our teaching force. We’re not going to get it with class size alone.

Question from Erin Wilcox, Principal, Colorado Springs Christian Middle School:
Does the research indicate an optimal class size for middle schools?

Douglas N. Harris:
We know more about the effects of class size in elementary schools because of large scale experiments such as Tennessee STAR and Wisconsin SAGE. I’m not aware of any class size experiments in middle or high school, nor do I hear many concerns from teachers at those levels.

Your question also raises interesting larger issues about WHY class size might matter. I’ve suggested in earlier responses that it is about classroom disruptions and differentiated instruction. It may also be that small educational settings create a more cohesive educational environment, where students feel a closer and better connection with classmates and teachers. This would explain why teacher aides do not seem to matter--since aides do not influence the “size” of the learning environment.

This is all somewhat speculative. I’m not sure researchers(at least not this researcher) know for sure why class size matters at the elementary level. Also, I would emphasize again that the test score effects are probably modest.

Question from Joshua Pierce, Research & Development, Edison Schools Inc.:
There is a great deal of public and professional support for reducing class sizes, however, the only research consistently referenced is Tennessee’s Project STAR and these results would be difficult for smaller districts and individual choice schools (e.g., charters) to replicate. Is anyone testing, or to your knowledge open to testing, the concept of moderately increasing class sizes, along the lines suggested by Mr. Cooperman, in conjunction with increasing compensation to access higher quality teachers (as well as possibly using some of the extra funds to invest in technology platforms to alleviate some of the administrative strain created by larger class sizes)?

Nancy Flanagan:
Thanks for asking this question. It represents some new thinking about solutions to optimizing class size, rather than old two-dimensional arguments. I have been involved in a new project out of the Center for Teaching Quality, called TeacherSolutions, where a team of accomplished practitioners is looking at new models in pay for performance. One of our ideas is that we could increase compensation for teachers who have demonstrated the capacity for excellence and willingness to take on larger classes, using flexible schedules and technology to effectively manage larger student loads. Why should truly exemplary teachers be limited to caps, when their teaching skill or engaging presentations could be more widely dispersed. A pilot for some of these ideas is long overdue.

Question from Veronica Martinez-Cantu, Graduate Student in Sociology of Education at University of North Texas:
Is increasing the classroom size and teacher salary enough to ensure all students will succeed? Is our goal as educators to teach students by the masses and hope they all learn to read and write or are we trying to produce students who can further our economy?

Nancy Flanagan:
I think you’re asking two separate questions here, Veronica. Increasing teacher salaries might be the right thing to do, but will not ensure student success. An economist will tell you that the right incentives improve performance, but research indicates that what teachers want most is support for their work, ongoing professional learning, and good working conditions (including collaboration and strong leadership). If teachers are to be paid more, they should be demonstrating strong knowledge and skills, special competencies, and a willingness to provide leadership (mentoring, coaching, professional development) to colleagues.

Class size is something else. There is some research that ties modest gains in learning to smaller class size (and other research that shows that class size doesn’t matter much at all). Perhaps policy-makers should stop assuming that leveraging high test scores will, as you say, further our economy. Taking the focus off which policies (class size, salary, teacher qualifications, etc.) get us higher scores and concentrating instead on what our students will need to know to succeed in a VERY different global economic environment might lead us away from forced, false choices in policy and toward different ways of thinking about how to better prepare our students for the road ahead.

Question from Jason Wermers, Reporter, The News-Press (Fort Myers, Fla.):
Has there been any study of the effects, if any, of Florida’s classroom size reduction amendment on student achievement? if so, what are the indications so far?

Douglas N. Harris:
I don’t believe that any results have been released about the effects of the Florida program, although I believe there may be a study under way.

It is important to emphasize that the Florida program applies in all school grades, K-12. Given the lack of evidence about the effects of small classes in middle and high school, this is almost certainly not a cost-effective approach. There was at one point some talk of a change that would narrow the program to elementary schools and shift the other resources to teacher salaries, but this appears unlikely--small classes are extremely popular with parents, teachers, and voters. (I’ve written elsewhere about why small classes are so popular and perhaps this will come up in the questions here.)

Question from Professor Cornelius Riordan:
Is there a What Works Clearinghouse or Campbell Collaboration meta analysis of the class size issue. If so, where is the best one or two such reviews located? And what what reported? Can this be provided during the presntation or sent to me so that I might obtain a copy?

Douglas N. Harris:
Almost all of the research has focused on the Tennessee STAR and Wisconsin SAGE because they are large-scale random-assignment experiments. I’ve summarized some of this debate in two articles (see citations below). The punchline is that the results are consistent across these and other studies, even though each study has its flaws.

There has also been a more extensive debate published by the Economic Policy Institute in Washington, DC, which is available for purchase online at the organization’s web site. The debate involves two economists, Eric Hanushek and Alan Krueger, who have studied the topic extensively. There is an older meta-analysis by Glass and Smith, which is also cited below.

Glass, G.V. & Smith, M.L. (1979) “Meta-Analysis of Research on Class Size and Achievement,” Educational Evaluation and Policy Analysis, Vol. 1, No. 1, pp. 2-16.

Harris, D. (2002). “Identifying optimal class sizes and teacher salaries,” In Cost Effectiveness Analysis in Education, (eds.) Henry Levin and Patrick McKewan. Larchmont, NY: American Education Finance Association.

Harris, D. and Plank, D. (2000). “Cost effective policies for reducing class size and increasing teacher quality,” in Allocating School Resources to Improve Student Performance. Chicago: U.S. Department of Education (North Central Regional Education Laboratory).

Question from :
With so many students being diagnosed (labeled?) with ADD & ADHD, active engagement,rather than passive “absorption” of information, has been the most effective teaching style. Doesn’t a small seminar-style class work best in terms of providing the most inclusive learning opportunity?

Nancy Flanagan:
I think we have to be careful with words like “best” and do more than an armchair analysis of teaching students with ADHS. An earlier question today proclaimed that “quality” students preferred large classes! We all have learning preferences, often developed through our personal success with a particular style. It’s harder to keep track of individual student responses in a large group, but I don’t think that’s a rationale for prescribing small groups for kids with attention deficit issues, necessarily. Perhaps we should ensure that our students are exposed to a variety of learning settings and instructional models.

Question from John Norton, education writer:
Some argue that initiatives to reduce class size drain the teacher supply pool and make it even more difficult for high-needs schools to attract well-qualified teachers. Don’t we need to deal with the teacher supply problem first?

Douglas N. Harris:
John, you raise a very important issue. There is more and more evidence that the quality of the teacher is central and reducing class sizes requires hiring more teachers. Moreover, because it is costly, it requires cutting teacher salaries, exacerbating the problem. While it’s difficult to say what class size is optimal, my research suggests that we’ve probably gone a little too far awith small classes on the average (though not necessarily in specific schools) and that we need to pay greater attention to the teacher quality issue.

Question from Saundra Koontz, teacher, IPS:
What are your feelings about the Tennessee Class size study (STAR)

Douglas N. Harris:
STAR is impressive because of the large number of students and use of random assignment. Nevertheless, there have been important questions raised about how the experiment worked in practice. As a result, I would say that the estimated effects are probably smaller than those reported, but stil positive. This is further reinforced by the evidence from other experimental and near-experimental studies (see my other web chat responses for more details).

It’s also important to reemphasize that even if the STAR are effects are real, “effective” does not mean “cost-effective.”

Question from Tom Wiedenman, Director of Teaching and Learning, Teaneck Public Schools, New Jersey:
Would you agree that class size should be incrementally decreased as you move to the lower grades with well under 20 students in Kindergarten?

Nancy Flanagan:
Based on the information we have, reducing class sizes for primary grades would seem to be one reasonable strategy for getting students off to a good start. Combining small class sizes with improved instruction would be even better.

Question from Lisa Andresen,: AmeriCorps, Summit Elementary and licensed 6-12 English educator ,:
What emperical evidence has been found that directly links SAGE and increased academic performance? and...what evidence substanciates the evidence that SAGE alone is responsible for this improved performance?

Douglas N. Harris:
SAGE is a program started in Wisconsin that includes a combination of class size reduction and other reforms. Students were randomly assigned to participate, which is why SAGE, as well as Tennessee STAR, receive so much attention and why there is reason to think the effects are genuine.

One question regarding SAGE the effects of SAGE is whether the effects on test scores were due to the class size as opposed to the other parts of the reform package. At least one study (by Molnar and colleagues) concludes that the effects of the other reforms were probably small and that the effect is mainly attributable to class size, although the effects are difficult to isolate.

Finally, let me note that the estimated effects in SAGE are similar to that in STAR, as well as to the average effect reported by Glass and Smith. If you’re interested in these studies, and my discussion of them, see the citations I put in one of the other questions in the web chat.

Question from Mike, High School English Teacher:
Mixed research notwithstanding, doesn’t it make intuitive sense that smaller class sizes should translate into “better” learning? It seems that the existing research has defined “better” as increased student achievement. Is it possible that “better” might fall into one of those intangible areas including a richer, warmer, calmer, deeper learning community and a more confident, passionate learner?

Poll 100 parents and ask them if they’d prefer a 35:1 or a 10:1 student-teacher ratio for their child, and it’s not hard to predict their response.

If it doesn’t seem to matter, then why would colleges and universities tout smaller teacher-student ratios as a selling point?

I just moved from 1st semester classes of 22 or 23 to 2nd semester classes of 30. It appears to matter at several levels. Is it possible the research isn’t asking the right questions?


Douglas N. Harris:
Mike, you raise an important point about the importance of outcomes other than achievement. The graduation rate, and even longer term outcomes such as civic participation, are two such examples that some studies look at. Unfortunately, it’s difficult to track students over such a along period of time, making it hard to determine whether class size influences those outcomes.

It’s worth noting the example of the Perry Preschool project which did track students through school, higher education and on into the workforce. Further, the effects of preschool on these longer-term benefits appeared were substantial. While we don’t know whether this is true with class size, there is reason to think that it might.

Question from Stephen Bartram, Rancho Buena Vista High School, Science Teacher:
How will reduced class size affect students who are disengaged from the curriculum? The issue is the student not the teacher or the curriculum or how great and grand the link between pedagogy and curriculum. What about the students lack of moticvation to learn?

Douglas N. Harris:
I think one of the greatest potential advantages of small classes is that students may become more engaged, for two reasons: (1) teachers may be better able to deliver more differentiated curriculum; and (2) the student may feel a closer connection to the teacher and other students, which may also motivate them. Put differently, the lack of student motivation may not be caused by the teacher per se, but smaller class sizes may allow the teacher to better address the lack of motivation.

Question from Todd Lacher, education reform research consultant:
Rarely does a single reform strategy, like class-size reduction, even begin to address the needs of a particular school. Reforms like this are usually much more (or sometimes only) effective when carefully paired with very specific mutually supportive structures/practices. Correct me if I’m wrong, but the crux of a reduction in class-size is the hope that by lowering the student-teacher ratio, the teacher will be able to provide a more personalized learning environment and more individually tailored instruction. It seems then that the conversation and research around the success and/or failure of class-size reduction should be heavily focused on defining what other structures/practices are needed to work in tandem with this strategy. From your experience/research what are the other reform strategies that work well with class-size reduction in creating a mutually supportive framework?

Nancy Flanagan:
Well--one model to look at would be the SAGE research, out of Wisconsin, which paired smaller class sizes with rigorous curriculum, accountability, professional development for teachers, and an extended day, all of which could well have a cumulative positive impact on student learning. I completely agree that it would be foolhardy to expect simple reduction in class size to give us uniformly great results. If we’re going to invest in an expensive reform like class size reduction, we ought to pile on as many potentially positive strategies as feasible, to maximize the investment.

Question from Michael Fixler, veteran (28 year) elementary school teacher, Ramsdell Elementary School:
Isn’t it true that keeping class size down for lower elementary students has been shown to raise achievement and is it not true that there are so many intangible benefits from lower class sizes in the younger grades, (e.g. more specialized attention to meet needs of students, better distribution of behavior problems, more time for teachers to work with individual students, more space in the classroom for students-with technology we are filling up our classrooms more and more, etc. the list goes on.)

Nancy Flanagan:
I didn’t get halfway through your question before I knew you were a long-time elementary teacher. Thanks for hanging in there, friend. You answered your own question, of course. I guess my challenge to you, as a veteran, would be to start speaking from your informed experience. I find it amazing that policy-makers sometimes put more trust in studies, when the overwhelming opinion of teachers, parents and students is that small classes work better. It’s easy to dismiss teacher pleas for smaller classes as whining about extra work. We need good teachers (assisted by parents and students) to provide credible information about the measurable benefits of reasonable class sizes to people who can make change.

Question from Geoffrey Enriquez, Teacher, IS204 Queens, NY:
Addressing class-size issues always seem to be avoided when educational issues are mentioned in the media. Are there any mechanisms or models anywhere in the country that enforce smaller classrooms (20-24 students) for general education students? And because smaller classes may not be economically feasible, especially in huge systems such as the one here in NYC, what are other educational factors that we should be addressing?

Nancy Flanagan:
Sure. Lots of class size management happens through teacher contract negotiations. Class size is a “quality” issue--in reducing class size, we have some backup for the idea that this might result in better student learning (compared, for example, to paying teachers more for graduate hours).

I think your second point is key, however. When smaller sizes aren’t feasible, there are lots of other things to try--a coherent alignment of curriculum and instruction, more time in school, teacher development, etc. No school should put all its improvement goals in one reform basket.

Question from cristina miclat, high school principa, philippinesl:
how can teaching be more effective if there are more than 75 students in a class?

Nancy Flanagan:
I notice that this question comes from the Philippines, so I’m assuming that this is, in fact, a personal issue for the questioner. I have very little experience in international/comparative education--but have seen amazing footage from South Africa (shot by an exemplary American teacher in the last decade) of extremely large groups of students, sitting on benches, cheerfully reciting and sharing chalkboards. “Classroom management” was not an issue, to say the least. All of this speaks to cultural expectations and the social value put on education. Our discussion today is on class size as it plays out in a particular environment. Children can learn from listening to a speaker with 500 other students, and there are some exceptional educators whose presentations should be dispersed to audiences of all sizes. I’m hoping that the dilemma of classes with 75 students will be solved through technology.

Question from Mark-Alan Whittle, CEO Street Advisor Canada:
My disabled son Logan recently passed away, but he was in a regular Public School class here in Canada, in the Province of Ontario.

We have a cap on class sizes up to grade three wherein classes have no more than 20 students, yet students learning ability has not improved.

If teachers were trained better would students not learn better automatically?

That is the real question here, how can all those students be hard to teach, it makes no sense. Teachers should be tested too, but they aren’t, masking over their failure to improve their ability to teach to all the children, including my disabled boy.

Thank-you for your candor and honesty if you dare to take this question on.

Nancy Flanagan:
First, Mark, my sincere sympathy for your devastating loss. It puts your anger and frustration in a unique framework. There is absolutely nothing worse than knowing your child is not getting what he needs to make progress. I think you’ve hit the nail on the head: some teachers do not have the training (or, if we’re being brutally honest, the disposition or the will) to be effective.

I can’t speak to Canadian policy on teacher testing, but it’s nearly universal here in the states for preservice teachers, and under NCLB many veterans have also had to be tested--and the research linking teacher test scores to higher student achievement and satisfaction is pretty mixed. I’m not sure that even regular testing will identify teachers whose work with students is substandard. We’d get better data from principal and peer evaluations, student achievement indicators--and from parents, like you.

Your comments go back, once more, to the core issue on class size: teacher quality is more important than teacher/student ratio. To pass class size legislation without paying attention to effective teaching in every classroom doesn’t make sense. It’s another silver bullet solution.

I wish you healing, and urge you to continue to monitor education policy and speak out.

Question from Joy White, teacher, Lee Hall Elementary:
Since it has been proven that smaller classes are better for teachers and students, why do we let them get so big? I think a class should never have more than twenty kids in any grade. I think test scores would improve if we kept them small and students would be able to get more individual attention.

Nancy Flanagan:
Why are classes so large? Because larger classes are cheaper. It’s a resource allocation issue. The STAR Project,the SAGE reseach and other studies were designed to examine whether smaller class sizes were a good investment in student learning. There is some evidence, as you noted, that small classes have a positive effect on student learning. But the key question is what percentage of available resources should we dedicate to reducing class size? Are there other things that raise student achievement as much or more, at lower cost? I know (and support) what good teachers think about the issue--we can serve kids better in smaller classes. Teachers need to develop their own research-based case for smaller class sizes and get into the policy discussion.

Question from Kalisha T. Miller, Education Program Specialist:
Having a special education background, I realize how important class size is for students with disabilities. However, I also know that without a good teacher, it does not matter if you have two students in the classroom. My question is, what other educational factors should be important other than class size?

Nancy Flanagan:
There are a vast number of factors that contribute to teacher effectiveness. With your expertise in special education, you have more credibility in identifying those markers in those who teach kids with learning disabilities than I, Kalisha. This does raise the point that it is important to learn from what we do know about effective teaching--coherent systems of instruction, paying attention to results, persistence, knowledge of students, content expertise, management skills, reflection, collaboration and on and on. The better we can define and replicate effective teaching, the stronger our case for reasonable class size.

Question from Martin Levine BOE member:
Parts of this discussion are a bit vague. What is “small”? We have class sizes that average about 20 in elementary and middle, but a bit higher in HS. Is that “small”? or is it medium?

Nancy Flanagan:
Adding to the vagueness, class size is often confounded with teacher/pupil ratio, a figure that is skewed by tiny special education classes, or incorporating other professional staff members (media specialists, counselors) into the equation. I think “small” is a relative descriptor; trying to achieve a figure below a state or national average doesn’t take into account contextual factors. If class sizes of 20 are working well for you in elementary and middle school, then you may well want to direct your resources toward other improvement strategies.

Question from Dean Leeper, Student, Georgia State University:
All of your research basically compares classes of around 25 to classes of around 15. Yet you call this “THE class size question”. What about classes of 8, 6, or 4? There is some research--Glass and Smith 1982, for example--that suggests you do not see any real benefits to class size until the class size becomes lower than 8 students or so. In other words, your discussion could be analogous to comparing French Fries and Onion Rings and calling it “The Nutrition Question: What Works Best?” I think the title of your chat is misleading and that you should call it what it is: a debate over reducing class size from 25 to 15 or not. Your response?

Nancy Flanagan:
I think the framework for this discussion is best and most efficient use of available resources. Our essential question is around what we know about real benefits of class size reduction, and how to balance those against competing needs. Your comments about the title of the chat do raise one point: research does show that tiny reductions in class size (i.e., 26 to 24) don’t give us the demonstrable boosts in student achievement. You do have to go from, say, 25 to 15 to get the big bump in scores. And speaking as a 30-year veteran, the teachers in my last (elementary) school would have been ecstatic with a guaranteed 15-student class limit.

Question from Susan Stryker, Kindergarten Teacher, Santa Clara Unified School District:
For the past two years, we have increased our class size in Kindergarten from 20 students to 30, due to budget cuts. We are noticing a difference in where our students are academically and socially as compared to our 20:1 classes. In addition, the First Grade teachers report that the students are coming in less prepared than in previous years. My question is, what is used to measure the success of students in smaller vs. larger class sizes and does it make more of a dfference in particular age/grade levels than others? If we were to choose a grade level where smaller class size is most effective, is there one that stands out?

Nancy Flanagan:
Choose a grade level where small class size is more effective? Sure. Yours. The STAR study was clear on this one: you get the greatest effect from small class size in the early grades. On the face of it, it makes sense. If there was any developmental level where more attention to socialization, strong development of learning foundations and reinforcement of good habits was important, it would be the primary grades. Have you thought of collecting and analyzing this data and making a presentation to your local policy makers? What a great example of teacher leadership.

Question from Stephen Bartram:
What is the relationship of class size to student DISENGAGEMENT? Will smaller class size address the issue of performance level of those students who are opting out of the curriculum?

Nancy Flanagan:
Survey research (asking teachers and students about their experiences in small/large classes) indicates that students feel they are more more engaged in smaller groups. A frequently read phrase: there is no place to hide in a smaller group.

I don’t think we can imply causality here, though. Smaller groups don’t necessarily lead to more engagement, attention or success. That depends on the instruction and interactions between teacher and student.

Question from Kathy Brown, parent:
Shouldn’t we assess the children’s cognitive abilities to determine if they require more or less instruction and interaction from the teacher? Also, I curious as to the number of edcuated mentors, or collaborative school community partners that are welcomed into the schools to offer one on one attention to students who may need some extra help. Programs like Big Brother and Big Sisters

Nancy Flanagan:
Your first question makes me a little uncomfortable. Although we do screen students for learning disabilities, and sometimes provide smaller learning settings, I’m not sure I would want to be part of a system where all students were assigned to classes sized by their abilities. Besides being a gigantic can of worms, all students can benefit from individual attention, as well as exposure to learning delivered in larger groups.

Your second question is a good one--why don’t we use more educated mentors to provide individual assistance when students are in large groups? I suspect the answer lies in the need to ensure that “learning assistance” volunteers are properly trained and doing productive things with kids.

Question from Jeff Johnson, teacher, South Division High School, Milwaukee Public Schools:
Good afternoon - I’m interested in reading your response to the class size question in light of the changing paradigm of classroom teaching as a result of technology - that is, the teacher no longer as the talking head spewing facts, using worksheets to assess a student’s comprehension, but engaging the student in various exercises and means to inculcate and evaluate a student’s analysis and synthesis of information. I guess the short of it is, how do the two of you see technology and its implications changing the classroom dynamic, thereby possibly affecting class size?

Nancy Flanagan:
What a great question. I actually believe that new ways of learning made possible through emerging technologies will bust open the whole issue of uniform class size (as well as any number of other instructional design and delivery issues). Distance and on-line learning, the technological “grandparents,” have already reshaped the concept of the bricks-mortar classrooms and access. This isn’t something we will be able to control, as innovations wash into our educational settings. I personally find the ability to deliver engaging educational experiences worldwide exhiliarating and look forward to the day when “class size” is an antiquated concept.

Question from Paula Petterson, Science Teacher, Ridgeview Classical Schools:
Can you please address the issues of class size specifically concerning science classes?

Nancy Flanagan:
I’m assuming you’re thinking about the desirability of smaller class sizes for labs and other hands-on work. Wouldn’t it be great if you could do whole class work with one sized group and have smaller groups for the labs--tailoring class sizes for efficiency as well as effectiveness? I think we could make a case for optimum class sizes and meeting times for all MS/HS subjects, but if we’re locked into the same class size for all subjects, we don’t get that flexibility.

I haven’t read any research that specifically targets science as a subject where smaller class sizes are ideal, although research does suggest that one-number blanket class size language is not the most efficient way to organize schools.

Question from Hayley Thompson, Education Writer, Ntl Cntr for Missing & Exploited Children:
What is the average class size in a public school setting? What is the average class size in a private school setting?

Douglas N. Harris:
The pupil-teacher ratio is quite similar in public and private schools--15.5 teachers per student. The average masks a lot of important variation; elementary school class sizes are smaller than high schools, for example.

Class sizes tend to be larger than pupil-teacher ratios because not all teachers are teaching at any given moment during the school day. They are also harder to measure for a variety of reasons that are probably not worth going into. Nevertheless, these reasons explain why we often have to talk in terms of the pupil-teacher ratio instead of class size per se.

A rule of thumb I established in one of my previous studies suggests that the class size is about 25 percent larger than the pupil-teacher ratio, which would put the actual average class size at about 19.4.

Question from Deanna Harris, Library Media Coordinator, Wake County Public Schools:
As an educator, I hear many teachers saying “If I just had smaller classes...”, “if they wouldn’t overload my classes...”. Do you all think it’s a classroom management issue (behavior) or is it a curriculum and instruction management issue (differentiation)? No matter the class size, my feelings are that good instruction -- differentiated -- reaches every student, helping him achieve. How can we help educators feel comfortable with the multi-ability classroom?

Nancy Flanagan:
Great question, Deanna. I have been trying here to fairly represent the opinions (and personal action research) of accomplished teachers who testify that class size does matter--that they can serve kids better in smaller groups. Notice that I said “acccomplished” teachers--teachers who constantly strive to improve their practice. For these teachers, the challenge is in differentiating learning and providing enough feedback for students to learn.

Some teachers are looking for lighter loads in the belief that smaller classes are easier, in terms of behavior management as well as multiple abilities. An economist might call this preference (for fewer responsibilities at the same salary) rational.

This is not research, but my belief is that most teachers go into teaching because they want to make a difference--a tremendous resource for schools and this nation. When we squander that resource by making working conditions in schools nearly impossible, we are needlessly wasting a great gift. I agree that good instruction reaches every student, at some level, and good teachers persist, persist, persist. I hope we can give these good teachers some support through reasonable expectations and class sizes, and pay them respect by listening to their voices.

Question from Garnett Arnold:
Based on my experience as a high school teacher, 12-15 students per class is about the right size, not from a teacher workload perspective, but from the individual attention and continuous assessment that is necessary in an environment where kids are just showing up for seat time, and hope to remain invisible. What does the research show?

Nancy Flanagan:
Wow. I can imagine HS English teachers across the nation rejoicing at the very idea of having a maximum student load of 15 students per class. The writing feedback they could give! The discussions they could lead! Nobody, as you note, could remain invisible. Ted Sizer suggests an optimum maximum teacher load for HS of 80. It’s not so much class size--the number you have at one time--as the total responsibility of responding to student work, and that applies to all subjects. The plain fact is that you can--if you’re doing your job well--spend more time providing feedback on student work when you have fewer students.

However--one of the problems with blanket pronouncements about issues like class size is that one classroom is not like another classroom. No researcher would (or should) prescribe an optimum number of students for best measured learning results, and I would discount any study proclaiming a best or most efficient class size at the HS level. (The STAR study gives us some guidelines with lower elementary classes.)

Tom Guskey’s work tells us that the optimum “class size” is one. Individual tutoring provides the maximum measurable learning results--but individual instruction (aside from being impractical at a school level) does not provide benefits of group learning: interaction, multiple perspectives, collaboration construction of new knowledge, etc. So it becomes, rightfully, an issue to be best settled in context--in order to advocate for an efficient class size, we need to know who, where and what you’re teaching, and what goals we expect you to reach.

My ideal class size was around 65 (I am a MS/HS band director). I needed the larger numbers to achieve a balanced instrumentation and a full sound. I developed processes to handle the large numbers efficiently, and the saving grace of my practice was that I had students for 4 or more years running, and got to know them and their learning styles well. I would never, however, declare--on the basis of my experience and preference--that a teacher who felt she was doing her best work with smaller numbers was wrong.

Question from Debbie Furtado, First grade teacher, Massachusetts:
My question is about how that an excellent teacher scaffolds her teaching in a regular size classroom to challenge a wide range of students and at the wame time develop accountability on the part of children to experience higher level thinking?

Nancy Flanagan:
The answers to your questions lie in books, not in a paragraph in an EdWeek on-line chat! But thanks for asking it--it does address the challenges of teaching in “regular sized” classrooms. No matter how large or small our class sizes, we all have the responsibility to push children as far as we can.

Question from Michael Bertram, President BOE, Denville, NJ:
What effect does the number of “included” students (IEP and 504) have on the optimum class size? Does research show a recommended ratio or maximum?

Nancy Flanagan:
The answer to your question is, of course, very complex, depending on many variables--the nature of students’ disabilities, the severity of those disabilities, the typical nature of instruction in the class, etc. As you’re aware, most weighting formulas are determined by the bargaining process in collective bargaining states.

Although determining best placements both for Special Education students and for effective instruction can be a tangle of frustration, determining optimum class size probably should be an “on the ground” decision, rather than something regulated by state/federal legislation or even research. The principles of inclusion--least restrictive environment, individually determined and flexible placements--are best made by people who know the “included” child.

Question from Emily Midouhas, Researcher, Northeast Islands Regional Education Laboratory at Education Development Center:
What is the impact of class size on student achievement for high poverty students? Does it vary by content area or grade?

Nancy Flanagan:
A rule of thumb on class size research is that you get the biggest bang for your reduction buck with very young children, as well as minority and disadvantaged children. The greatest benefits to students come in the first year of placement in a small class. I have also seen studies that suggest the largest gains from small classes are made in literacy.

I think those are good arguments for directing resources in those directions. A better policy, however, would be combining reduced class size with improved instruction, coherent curriculum, and frequent assessment. If resources are directed to class size reduction, we ought to maximize the benefits.

Reducing class size for certain students and certain subjects is a more efficient way to use scarce resources. It might be a hard sell to teachers (class size reductions for certain grade levels or student groups) who are looking for straight equality rather than efficiency or equity. Using research to target resources where they are likely to do the most good makes sense, however.

Question from George Scheuermann, part-time faculty, University of Dayton:
As many districts struggle with teacher recruitment and retention, what might the impact of class size (and the added work load)have on teacher morale, burnout, absenteeism and enthusiasm?

Nancy Flanagan:
This is a great question. First--it’s important to clearly distinguish between class size and teacher workload. If teachers are assigned more, but smaller, classes and their overall student load does not decrease (or increases), it’s highly unlikely that they will feel at all positive about class size reduction, even if their classes are quieter and more personal.

One of the problems with rigid district- or contract-mandated parameters on class size is that teachers have little say over their optimum instructional delivery model. A U.S. History teacher, for example, might prefer to deliver lectures or supervise small-group work with 45 students at a time leaving two additional teaching periods for reading student work and conferencing. A kindergarten teacher wants the smallest class possible, because her work involves building strong learning habits and classroom community. One model is not optimal for everyone. Lockstep class size language does not permit this kind of flexibility.

Richard Ingersoll tells us that teacher satisfaction, enthusiasm and willingness to remain in a particular school are very dependent on teachers’ perception of their own influence over “social” work issues: tracking, handling student behavior problems and genuine input into the numerous policies necessary in effectively running a school building.

Perhaps if teachers--within the bounds of best practice in school scheduling--were given more latitude in deciding how to best use their skills to determine class size, we’d see less burnout and absenteeism.

Question from Janice Zmrazek, SAGE Program Coordinator, Wisconsin Department of Public Instruction:
Doesn’t the effective schools movement tell us that there is no single strategy that will work well in every situation and location - but instead suggest need for a comprehensive package of initiatives faithfully implemented? Small class sizes may be the one means that makes time for many of the other effective school components (participation in collaborative decision making, working with parents, participation in professional development, individualized instruction) and also serves to improve working conditions and morale, thereby encouraging good teachers to stay in high-need schools.

Nancy Flanagan:
Your comprehensive question provides its own answer--not surprising since you’re associated with one of the more comprehensive and well-designed class size research projects.

The SAGE (Student Achievement Guarantee in Education) study is unique in that class size reductions were coupled with rigorous curriculum, extended days, strong accountability measures and aligned professional development. The early results showed students in smaller classes making significant gains, measured against similar comparison schools, with the results most impressive in minority students.

This is a good opportunity to reiterate that class size reduction alone has a much more limited impact than smaller classes coupled with improved instruction. The California class size initiative is a good case study on simply reducing class size, without accompanying attention to instruction, teacher quality or even adequate and appropriate physical space for the additional classrooms required. Fewer kids with a marginal teacher in a converted closet doesn’t raise student achievement. Class reduction is an easy (albeit very expensive) “solution” to a problem. Defining effective teaching, and looking at context to determine optimum class size is much more nuanced and difficult. It’s no surprise that policy-makers turned to the quick fix.

Question from Cynthia Pugh-Carter, Student of Education, Grand Canyon University:
Research suggests that lowering class sizes does not necessarily lead to higher student achievement. I believe it is a start in the right direction, having experienced the chaos of trying to teach 33 students. What is your position?

Nancy Flanagan:
Chaos, indeed. While simple econometrics find little evidence that smaller class size is firmly linked to higher test scores, the overwhelming opinion of teachers, parents and students is that classes can be too large, and kids can fall through the cracks. Why the “common knowledge” of consumers and providers of education--that class size matters, a lot--is ignored by economists is a good question.

Economists are looking for efficiency, getting maximum quantitative results at the lowest cost. Parents, students and teachers are looking for something else: attention and more personalized instruction.

Teachers and students say that smaller classes have fewer distractions, lower noise levels and more space, in addition to a greater likelihood of differentiated instruction. Teachers track student learning and student behavior more effectively in a smaller group.

The salient fact: class size reduction is expensive. It may well be worth it, especially for our youngest children (3rd grade and under) and for poor and minority children. Doing a cost-benefit analysis on provision of education services for the poor and very young children, however, involves a kind of steely perspective that probably comes more naturally to an economist than a teacher. These analyses must be done, however; every worthy educational initiative, including reducing class size, involves tradeoffs. It’s important to have good information on the most effective allocation of resources.

Question from Jill Andersen, 2nd grade teacher Sioux Falls, School district:
Our jobs are not to just simply teach. Everyday we diagnose, identify students with learning issues, analyze data, counsel and assess. I have 29 seven and eight year olds with many different needs. How effective would a doctor be if he had 29 patients at one time- doing all we do? How about a nurse with 29 patients taking 29 temps at a time? Can you imagine a pharmacist filling 29 prescriptions at one time? On top of all we do, we still need time- lots of time- to build relationships too. It is impossible to be effective and do what we do with 29 students. How can data show small class sizes aren’t effective? Please address this issue. Thank you

Nancy Flanagan:
Hi Jill. I think your heartfelt question illustrates one of the problems of quantitative research--it’s difficult to capture the realities of real children with real needs in large-scale data sets and statistical tools (even the newest, spiffiest statistical models).

We have to give some credence to teachers who tell us almost universally that they do a better job with smaller numbers of kids. We need to listen to our best, most accomplished practitioners talk--as you did--about the complex practice of teaching, and the layers of responsibility that teachers assume for their children. Children can’t learn until and unless their other, more basic needs are addressed, and reducing the class size question to an equation structured around standardized test scores and student/teacher ratios leaves out a lot of data that isn’t easily measured--things like the important relationships you mentioned. I’m sure any parent would choose “a teacher who cares for my child and is committed to their overall learning” over “a teacher who has the skills to raise test scores.”

I also think teachers should develop arguments for appropriate working conditions in education on educational merits. We are not doctors, lawyers or architects. Our work is unique and vitally important; we need to tell powerful stories from legitimate action research we do in our own classrooms. If we are constantly working toward the goal of more effective instruction, our own credible observations and data collection can be far more influential with policy-makers than “remote” statistical analysis.

Kevin Bushweller (Moderator):

Thank you for joining us for this informative discussion. And a special thanks to our guests for taking time out of their busy schedules to offer their insights on this important topic.

This chat is now over. A transcript of the discussion will be posted shortly on

The Fine Print

All questions are screened by an editor and the guest speaker prior to posting. A question is not displayed until it is answered by the guest speaker. Due to the volume of questions received, we cannot guarantee that all questions will be answered, or answered in the order of submission. Guests and hosts may decline to answer any questions. Concise questions are strongly encouraged.

Please be sure to include your name and affiliation when posting your question.’s Online Chat is an open forum where readers can participate in a give- and-take discussion with a variety of guests. reserves the right to condense or edit questions for clarity, but editing is kept to a minimum. Transcripts may also be reproduced in some form in our print edition. We do not correct errors in spelling, punctuation, etc. In addition, we remove statements that have the potential to be libelous or to slander someone.

Please read our privacy policy and user agreement if you have questions.

Chat Editors

Related Tags: