Class size does matter: at least for disadvantaged students

Larger class sizes can have negative impacts on disadvantaged students. www-audio-luci-store-it/Flickr, CC BY-SA

A recent report by the Victorian Competition & Efficiency Commission has suggested that reducing class sizes in the state has not improved student academic performance.

The report said that despite 50% of increases in funding Victoria’s teaching workforce over the past decade going towards reducing class sizes, the impact of smaller classes has been negligible. It suggested that increasing class sizes to the maximum levels would save around $300 million a year.

The reasonable conclusion to draw from this report is that smaller classes do not equal better learning environments.

But is this really the case?

Size does matter

My research has shown that class size has a significant effect on academic achievement. The impact of a small class is particularly strong in the early years of schooling, and it is especially important for students who are traditionally disadvantaged in education.

I recently completed a meta study of 112 papers on class size and academic achievement, and found that:

  • Small class sizes in the early grades are significantly beneficial for students, especially when a class is reduced to fewer than 20 pupils

  • These gains are particularly strong for disadvantaged and minority groups

  • The longer students are in small classes, the greater the benefits. However, even when students are returned to larger classes in later grades, the benefits of small early classes persist

  • The benefits can be seen in both tests of measured achievement and other measures of success

  • The evidence for the advantages of small classes in the upper grades and high school is so far inconclusive.

The VCEC report is based on research that cherry-picked from the evidence. Ed Ivanushkin/Flickr, CC BY-SA

Cherry-picking season

So why did the Victorian Competition & Efficiency Commission (VCEC) find that class size has no impact? The VCEC has cherry-picked from the evidence and used flawed research.

The report’s conclusion is based on claims and research by Dr Ben Jensen, previous education program director at the Grattan Institute.

I have written before about the problems with Dr Jensen’s report on Australian education and teacher quality.

The main issue is that the report was based on research by the conservative American economist Professor Eric Hanushek, which has now been widely discredited.

In a ruling on a school funding case based on Hanuchek’s findings, a Denver judge stated:

Dr Hanushek’s analysis […] contradicts testimony and documentary evidence from dozens of well-respected educators in the State, defies logic, and is statistically flawed.

But discredited evidence isn’t the only problem with the VCEC report. It also uses teacher-student ratios as a proxy for class size and state averages rather than maximum class sizes. This masks the fact that some larger classes across the state are being levelled out by smaller classes, particularly in regional areas.

The report also cites evidence from Singapore and Korea to support the claim that larger class sizes are possible while improving student achievement. My research suggests that these findings are not fully applicable in an Australian context.

Specific cultural factors are at play in Singapore and Korea, such as a Confucian respect for teachers, many hours spent doing homework and parents spending billions of dollars on private tutors. Moreover, teachers in these countries have much more time to prepare, give feedback and collaborate.

What should we do?

The VCEC report does include some excellent recommendations for action, such as enhanced principal autonomy.

It also finds that an individualised approach to class sizes is necessary, saying:

class size policies should be tailored to specific situations, where benefits are likely to be realised, rather than a ‘one-size-fits-all’ approach.

Given my findings that small class sizes have a significant impact in the early years and for disadvantaged students, but that evidence is inconclusive for older grades, I agree that a tailored approach is necessary.

In contrast to across-the-board class-size reductions that teacher unions have called for in Australia (and the rejection by government of such proposals as fiscally irresponsible), a targeted approach is a feasible way to make class-size reduction effective and affordable.

Reducing the size of classes does not need to happen in every subject at every grade level in every school. For example, class sizes could be reduced specifically for numeracy and literacy classes.

Using a combination of redeployment of existing staff and addition of special literacy and numeracy teachers, it would be theoretically possible to have small classes (average of 15 pupils) with a much lower additional cost.

To ensure that the schools that need small classes the most get the help they need, Australian public education needs a more nuanced funding program. Well-resourced schools (mainly schools in middle-class suburbs) do not necessarily need the smaller class sizes that disadvantaged schools require.

The class size debate should now be about weighing up the cost-benefit of class-size reductions and targeting the areas that need help the most. It is clear that increasing class sizes across a whole school system as the VCEC has suggested may harm student academic outcomes, in particular for students from disadvantaged backgrounds. Saving money now by increasing class sizes may result in substantial social and educational costs in the future.