Bigger Classes in Schools? Will they ever learn?

By Michael Edmonds 03/02/2012

Yet again the topic of class size in schools is in the news. A report from Treasury is suggesting that schools increase class sizes as a cost cutting measure for the government. It is disturbing to see it reported that Finance Minister Bill English has made comments that there is “clear evidence” that class size doesn’t matter. As someone who taught high school for a very short period of time about 18 years ago I can tell you from experience class size certainly does matter, particularly we we consider the many expectations that have been placed on teachers and schools over the years – all of the marking and the feedback, the expectation that schools provide sporting coaches and healthy eating advice to students, and most recently that teachers may be legally obliged to watch more carefully for abused children.

Some classes around New Zealand already exceed 35 students, which if you consider that on an individual basis equates to less that 1.5 minutes of teacher time per student. Treasury is recommending the educational equivalent of battery farming. We already have students who fall out of the system, do we really want more to follow?

If we seriously want our children to succeed there needs to be a push for smaller classes, not larger. In the long run, the benefits from the improved educational outcomes should outweigh the initial costs. (after the very good point made by Dave below I’m retracting this comment. In determining how to better serve the educational needs of our children, I think much more needs to be considered than just class size).

I think this cartoon from the Christchurch Press somes things up very nicely.

Larger classes in schools?
Larger classes in schools?

0 Responses to “Bigger Classes in Schools? Will they ever learn?”

  • Michael, I really enjoy your work debunking various questionable health practices where you take an evidence-based approach, but this post seemed to go against that. My understanding is that the research in this area show small positive gains to reductions in class size, although that is bound in part (as I recall) by particular ranges. Other practices, such as improved teacher professional development, have been shown to have a greater impact on student learning than reductions in class sizes.

    I’m not an expert in this area but I found this very interesting link exploring the effectiveness of different approaches to school student achievement and reducing class sizes was in the middle of the pack.

  • I’m on the fence about this. I gather that the existing research (
    is sparse and currently only supports increases in performance when class sizes are on the order of 15 students. As you’ve pointed out many class sizes are in the 30-35 student range. This means a reduction by half. With a concomitant requirement to higher twice as many teachers.
    With the job of teacher not being seen as particularly desirable can we really source that many extra teachers, even if there was the will and money to do so?

    It could be the education dollars might be more effectively spent on a different strategy. Perhaps focusing on the quality of the curriculum, better teacher training/mentoring or computer aided education would provide both increases in education quality and be more cost effective.

    Urging that these approaches be looked into as well might be more productive that a knee-jerk reaction to increasing class sizes.

    Perhaps our resident educators could provide a perspective on this?

  • What frustrates me with this emotive subject is that John Hattie’s meta-analysis is almost always trotted out as evidence that class size does not significantly impact learning. (Hattie’s Visible Learning research provides 105 more significant factors).

    Few seem to have read his research (possibly because it’s locked up in an evil elsevier journal). Hattie makes the point that class size can indeed make a difference, but teachers must first change their pedagogy. Unfortunately few teachers changed the way they taught despite having fewer students to teach. Expecting improved learning outcomes when teaching the same way is, of course, insane.

    Hattie would argue there are 105 other influences that have a bigger impact on student achievement and we should focus on those that make the biggest difference.

  • Hi Dave
    Improved teacher professional development – brilliant, I fully support that. But that does not mean that would allow greater class sizes to work.
    If you could pass on any further research references that show that reducing class sizes has little effect I’d appreciate it, but if memory serves me correct most studies that tend to be quoted are from countries with significant cultural differences.
    The only valid research would seem to me to have to be derived from looking at New Zealand schools and cover all deciles, including those where most classes have at least one child with some sort of learning disability.
    Also if you look at many of the factors listed on that table you linked to they require teacher time, for example 1 to 1 tutoring. It is hardly going be effective with less that 1.5 minutes per student. Same for homework – someone has to mark it.
    In considering changes to how we teach, student progress is not the only factor to consider. The mental well being of both the students and the teachers is another important factor to be considered

  • As one of the ‘resident educators’ I have to say I’m torn on this one. Yes, changes in pedagogy do have very positive effects – heck, I’ve written about that myself! However, it’s one thing to urge such changes & another to spend the time & money on professional development so that teachers can learn about those changes & take them on board. I would hate to see larger classes coming back in without that support for teachers. Especially when, as Michael says, there are so many other things expected of teachers in a day’s work.

  • If you go to google scholar and search you’ll find plenty of peer-reviewed evidence. I searched on “effective teaching biggest impact student achievement”. I also searched on “class size biggest impact student achievement” and got the same sort of results. Here’s one to proove they’re there:

    It is interesting to see Michael using so many of the techniques of the pro quack medicine brigade, like “I can tell you from experience” – never mind a peer-reviewed study. “Some classes around New Zealand already exceed 35 students, which if you consider that on an individual basis equates to less that 1.5 minutes of teacher time per student.” Some classes – what proportion, what subject? What is the mean, median, mode? “1.5 minutes of teacher time” – that’s roughly per hour (not stated) and assumes (not stated) all teaching is one to one.

  • Possum,
    So I shouldn’t mention my experience of teaching at all? At no point did I say my experience outweighs peer reviewed studies, which is what the pro quacks do. Furthermore, most studies do show that a reduction in class size does have a positive effect on teaching. It is unfortunate that the reference you provide only includes the first page, from which very little information can be gathered. Interesting that it has published in an economics journal and not an educational one.
    Granted the 35 plus figure was based on media reports from teachers representatives so it is difficult to know how many classes have this number of students but from my experience 35 students is a very large number of students to have to deal with. I used to teach around 24 and in a busy classroom that is more than enough to keep you busy.
    In the Hattie study often used by the government to support increased class size, it does show that reducing class size does have a positive benefit but smaller than other aspects, including the often quoted “quality of teaching”. However, if one looks at other aspects which promote student achievement they include tutoring, direct instruction, early intervention and lack of disruptive students – all of which can be assisted by smaller class sizes.
    Teaching and learning is a complex process – when politicians pick out one “fact” to supposedly support their argument, when in fact it does not, it frustrates me. My apologies if in my enthusiasm to counter this I perhaps didn’t include as much factual support as I could have.

  • Possum

    In 2011 class size for years 7 to 13

    Median 25
    Mode 28
    Mean 23.3
    8.4% classes over 30 students
    Maximum = 51 students

    From PPTA report

  • I think you make a good point Michael: teaching and learning is a complex process. By reducing teaching into a series of interventions we risk losing sight of the fact that many are interconnected. So, for example, the teacher-student relationship has, according to Hattie, 3.5 times the effect of class size on learning (almost twice the average effect), however, a positive relationship with students is far easier to achieve in smaller classes. It’s complex!

  • I don’t know which specific Hattie paper is being referred to, but in the freely available download of his inaugural professorial address in 1999 he states (page 8) “CLASS SIZE The research on the effects of class size has been among the more voluminous in educational research with very systematic findings:
    * Achievement, attitude, teacher morale, student satisfaction gains are appreciable in smaller classes, so long as we recognise that small classes mean 10-15, as there are negligible gains between 40 to 20 students per class.
    * This effect was the same for primary and secondary schools, across all subjects, and across various ability levels.
    * There is little evidence that instruction methods change when class size is reduced, although a large part of improvement can be explained by improvements in student task engagement.”

    This is not a not a single piece of research by Hattie, it is a summary of a whole lot of research by other people.

    The finding that class size is not very relevant to achievement finding also appears in the very comprehensive OECD PISA studies. Unfortunately the most recent (2009) PISA study is not available free on the web.

    Michael. Thanks for the extra information. When you state “Furthermore, most studies do show that a reduction in class size does have a positive effect on teaching” you need to be a lot more specific, in the light of the information above, such as a reduction from what to what. Also, please note with respect to your style of arguing that you queried the example I gave being published in an Economics journal. Any study of the impact on pupil outcomes of class size is going to involve a lot of statistical and mathematical analysis to account for socio-economic, gender and ethnic differences etc so that class size is the only variable left. An economist can do this just as well as anyone else and an economist will be interested in getting the best outcomes from the least resources.

    I think one thing we need to know more about is how one defines better student outcomes. PISA uses scores in reading, maths and science. I have no idea how anyone else defines outcomes and if we don’t know what we are looking for we won’t know when we’ve got it.

  • Stephen,
    I think you have put your finger on it, teaching and learning is a complex process where a lot of the factors are inter-related.

    Possum, you finish with a very good point – what do we define as better student outcomes? This should really be the starting point in deciding what approaches to teaching are used.

    Also, it occurred to me last night is what size of classes are our schools currently designed for? Most of the classes I’ve been into around Christchurch seem only really big enough to hold 30 maximum. If the government wanted to go higher than that, then the physical environment of NZ classrooms will need to change.

    Reading through the Hattie paper that is commonly referred to by those in favour of increasing class numbers (International Journal of Educational Research 43 (2005) 387 – 425 it really does reinforce the challenges of educational research for example how do you control the many variables that affect learning in the classroom setting? Also for many of the research papers Hattie reports other papers which debate what the authors have concluded.
    The paper includes a list of the various influences on student achievement. The top 10, in order of effect are:

    1) feedback (0.81)
    2) direct instruction (0.81)
    3) prior achievement (0.80)
    4) lack of disruptive students (0.79)
    5) quality of teaching (0.67)
    6) phonological awareness (0.66)
    7) early intervention (0.64)
    8) peer assessment (0.63)
    9) Challenging goals (0.59)
    10) Self assessment (0.56)

    But how many of these would work better in smaller classes?

    Possum, I’ve emailed you a copy of the Hattie paper, with your sharp insight perhaps you will let me know what you think of how well you think the research relates to teaching in NZ and how well variables have been controlled?

    If the NZ government is serious about increasing class numbers I would like to see some good educational research carried out in NZ classes to give us some meaningful data on what the effects are likely to be.

  • Thanks for the paper, Michael. I have read all 39 pages as best I can, given his writing style. Hattie is not trying to be descriptive but to find what works. He goes into quite a bit of depth analysing the papers whose results differ from the majority. Hattie is also very widely read in this field and quotes about 150 other papers in the 6 pages of references at the back. It is totally inappropriate for anyone to try to dismiss him on the grounds it is just one paper.

    Note that Hattie is not dismissive of small class sizes per se. In his paper on the page numbered 417 he states:The argument in this paper is that those teaching practices that are conducive to successful learning are more likely to occur in smaller rather than larger classes, and these practices do not actually occur more in smaller classes because teachers have been prepared to, and indeed do, work with larger classes using more transmission practices and therefore they are not so equipped to adopt the more effective practices when they are given smaller classes.

    Translating that quote into English, I infer that smaller class sizes may facilitate the various influences on achievement, but do not themselves produce increased achievement.

    Your final question above interests me. We know that for real sciences like physics, chemistry, geology, etc that a result anywhere in the world, is applicable anywhere. If we add sodium to chlorine we get salt, etc. I would be interested as to why you think oversea studies would not be applicable for NZ schools.

    Where I think we really do need study in NZ is to find a way to stop failing the bottom 25-30% of our children. The OECD PISA studies show consistently that NZ has the widest gap between the achievement of the top and bottom ranked students in any of the 30 or so countries studied. This is mainly a cultural effect, given that most of the failing students are Maori or Pasifika. So NZ study is needed.

  • Possum,
    Re the cultural comments, the actual management and discipline issues can vary from culture to culture. For example, the high value placed on conformity and structure in many Asian countries mean that less time is spent on disciplinary issues and more on teaching than in western countries. However, on reflection, given that many of the references are from Western countries this is probably not a huge difference. Still each country has different cultural backgrounds and attitudes to education so I would love to see some NZ based studies.
    You talk about the failings of the lower 25-30% of classes and instinctively I would suggest that smaller classes would be an obvious solution however, our previous discussions here suggest that more thought needs to be put into solving this problem. It would be interesting to know what has already been tried in NZ to solve these issues and what has and has not worked.
    I agree that (Prof?) John Hattie’s should not be dismissed as it is a comprehensive and interesting piece of work – what we should be cautious of is how it is interpreted, and be very wary of selective use of what it contains, whether that is by politicians or the occasional blogger. 🙂

  • I have to say that when I was a secondary teacher, items 1) feedback (0.81), 2) direct instruction (0.81); & 4) lack of disruptive students (0.79) were easier to achieve/less common in the smaller classes I taught. For 1 & 2 it was simply a matter of time 🙂 I’d have to say that meaningful early intervention was also a bit easier to achieve.
    Your comments on classroom size are very apt, Michael; most of the classrooms I’ve been in (& I see quite a few each year) really only cater for 30 or so. I remember having one class of 35 in a lab that had seating for 30; the ‘extras’ ended being crammed in at side benches, which wasn’t really satisfactory for anyone.
    Yes, I realise that’s anecdote!
    7) early intervention (0.64)
    8) peer assessment (0.63)
    9) Challenging goals (0.59)
    10) Self assessment (0.56)

  • Moving on from Hattie to your original blog, in the second sentence you suggested this was a cost cutting measure. It wasn’t. In the Treasury Briefing for the Incoming Minister there are recommendations (page 4) for Education:” Reform the education system to improve educational attainment at lower cost:
    – Target existing early childhood education funding to children in low-income households
    – Implement initiatives to improve school teacher quality, funded by consolidation of the school network and increasing student/teacher ratios
    – Reintroduce interest on student loans and target tertiary funding to younger students and higher level qualifications

    The idea was reprioritising funds from one area to another within education, not taking the funding away. Treasury blithely ignores the fact they are in practice suggesting a major year fight with the teacher unions to introduce performance pay, which has overtones of bulk funding.

    The detail from which the relevant two lins in the summary was drawn is on page 21.
    New Zealand’s compulsory education system produces good outcomes for most students, as evidenced by our strong
    performance in international tests. However, despite large funding increases, achievement levels remain unacceptably
    low for some groups. Student achievement can be raised by improving the quality of teaching, which the evidence shows is the largest inschool influence on student outcomes. Increasing student/teacher ratios, and consolidation of the school
    network, can free up funding that could be used to support initiatives to enhance the quality of teaching, such as
    more systemic use of value-add data and a more professionalised workforce.

    To be fair to you, this text does come under a heading referring to freeing up funding. But the freeing up funding part relates to reintroducing interest on student loans.

  • Possum,
    So, while the briefing is in fact cutting costs in the area of class size, what you are saying is that the money saved will be moved elsewhere in the education system? Hence, overall this should not be viewed as a cost cutting exercise, just a redirection of funding?

    Sounds reasonable to me. So long as the education budget at the end of the year is no lower that the previous year, then costs have not been cut.

    What do you think “value- add data” refers to, and do you tihnk a more professionalised workforce means they will be investing more in teacher training and professional development?

  • Yes, that’s what they are saying, just a redicrection. I don’t know what Treasury mean by value-add data, but I suspect they mean trying to take funding from areas lower on Hattie’s list and spending more in areas that are higher, which I think would probaly include professional develiopment etc. Nothing is going to happen in Education solely on Treasury’s advice, but I expect Ministry of Education advisers will be asked to comment.

  • Afterthought. All government department briefings to incoming ministers (BIMs) were released last week. You might want to have a look at what the Ministry of Education said to its Minister – all BIMs are on the relevant websites. As far as I saw, journalists only bothered to report on Treasury’s BIM.

  • I like your use of the word ‘inventing’, David.
    Anecdote warning: way way back when I had been teaching secondary school for a few years, the government of the time (it was National) first really got into talking about performance pay. The idea was to look at how a teacher’s classes did in School C etc, if I remember rightly. Those of us teaching the ‘remedial’ classes (those kids who were highly unlikely to achieve much in the inflexible SC system, although we could certainly measure improvements in the classroom) were just a tad concerned about this, as we’d score rather poorly…

    I do think, though, that some means of gaining value-added information is definitely necessary. At present lower-decile schools often get hammered in the NCEA rankings, & I do wonder how much of that is related to where the kids are at when they arrive at the school. This is not to diss the primary schools, because things like home environments can also have a big impact on achievement. So it could be that some schools do a heap of good work to raise students’ achievement from the point of entry, but without knowing what that point of entry is, they (the schools, & their teachers) maybe don’t get recognition for it.

    I will stick my neck out & say it’s a fair bet (& indeed there is evidence to support that bet) that many uni lecturers don’t really have much idea of where their new first-year students are at, which a) makes it hard to bridge them into the uni system & b) means we’ve got little to go on in terms of assessing how much difference we’ve actually made to their learning. Some form of ‘pre-teach’ assessment would be extremely useful here.

  • I’m with Michael Edmond here. The less number of students a teacher have to teach, the more time the teacher has to spend with each one.

    Possum said…
    If you go to google scholar and search you’ll find plenty of peer-reviewed evidence.

    Those who have done the peer-review have probably never done any real teaching at all.

    BTW, I’m not a teacher, never have been one (or have formal qualification as a teacher), but I have coached in the evening (in physics/maths) many pacific island students in the last 8 years or so and I have noted (anecdotal) that the less number of students I have, the more time that I spend going one on one with each student. Their understanding is lifted dramatically, when I spend more time (going one on one) with each. When I have 4 or more per coaching session, then I don’t have enough time to go one on one with each student.

    Sometimes when I asked the students if they understand a particular new topic that we have covered in a session and they all said yes (all the time), only for me to find out in the next session when I have time to go one on one, that 1 or 2 students didn’t really understand what we covered in the previous session. This is how I discovered holes in their knowledge and I patched them as I found them. You won’t discover those holes by just asking them directly if they follow/understand the topic, because sometimes they could be mistaken that they fully understood a certain topic but in reality they don’t and this is what I’ve seen all the time.

    A teacher IMO can only have a chance to spend more time with each of his/her students if the class size is small.

  • Said it before and I’ll say it again… The class size thing, while interesting, implies a change that involves a complete re-education of every teacher in the NZ education system. It also implies a change to the infrastructure as noted by Alison – 35 into 30 doesn’t go in anyone’s mathematics , even the convoluted econo-math of Treasury.

    But further, if you take a look at hte factors on Hattie’s list, several jump out – classroom disruption, preparedness for education, and a number of others slightly further down are primarily about home life and the ability of the student to integrate into learning seamlessly from home. This is demonstrably difficult when “home” is an under-resourced, health deprived, disfunctional result of 25 years of laise faire economic management.

    So, the best Treasury could do for education is to stick to its knitting (and some would add, quit dropping the stitches all the time) and focus on economic advice that addresses real equality of access – to education, health and society in general.

  • The starting point has to be the question: “What are schools actually for?”
    You can organise the system around a very few high-priority ‘student outcomes’ but you can’t optimise all of them. Do we want better outcomes for elite students? Do we want to fix the long tail of underacheivers? Do we want schools to fix social or family problems? Do we want to earn more money from foreign students? etc etc. The prioritisation is essentially a political question.
    As an aside, although a pertinent one, I have found that most criticism of schools and teachers comes from those who have never taught, and don’t intend to try.

  • As an aside, although a pertinent one, I have found that most criticism of schools and teachers comes from those who have never taught, and don’t intend to try.
    Heheh. Back in the day (& now too, actually) I was often the recipient of comments along the lines of ‘you school teachers, I don’t know what you’ve got to whinge about. Short working days & long holidays, you’re so much better off than everyone else.’ My usual response was/is, OK, so why don’t you take up teaching if it’s so easy & such a cushy number? The answer was almost invariably on the theme of ‘no way, you wouldn’t catch me in the classroom!’
    ’nuff said.

  • A several people have indicated education is a complex process and there are other factors than class size which can have a greater effect.
    For example I would rather teach a class of 30 well behaved students than a class of 15 where several students are disruptive. When I taught high school there were no effective ways of removing disruptive students which I found completely frustrating.
    It would be nice to see a re-examination of how we teach in NZ where teachers are widely consulted as well as relying on educational reports.