While teachers often insist that smaller class sizes allow for more effective learning, academic research provides little evidence to back this up.
But research often measures only test scores, whereas teachers also observe an improvement in pupil attitude and behaviour, according to Peter Blatchford, professor of psychology and education at the Institute of Education, University of London.
In a chapter entitled "Class size: is small better?", which appears in Bad Education by Philip Adey and Justin Dillon, Blatchford questions whether small class sizes are in fact best for pupils.
"What could be more obvious: fewer pupils in a class is surely better for the pupils and for the teacher," he begins. "One of the main reasons parents give for spending money on private education is that class sizes are smaller ... Teachers are also very strongly of the view that small classes make their job easier."
But, he adds, even small reductions in class sizes can be extremely expensive. As a result, "some politicians and policymakers worry that teachers' arguments in favour of small classes are more about making life easier for them ... than raising pupil performance".
He concludes: "The debate over class size is intensely political."
Blatchford examines existing research on class size: essentially, two influential studies. The first, involving more than 7,000 pupils in the US state of Tennessee, showed that pupils in small classes of up to 17 children performed significantly better than those in classes of between 22 and 25 children. These effects were still in evidence a year after pupils had returned to normal-sized classes.
The second piece of research was directed by Blatchford himself. More than 10,000 pupils were tracked through primary school. Even once other mitigating factors were accounted for, this revealed that class size had a clear effect on children's attainment in literacy and numeracy during the first year of school.
Effects on literacy progress were still evident by the end of Year 1. But by the end of Year 2 effects were uncertain. There was no effect on mathematics after reception year.
"The conclusion seems clear: for the youngest children in school, the smaller the class, the better," Blatchford says. But he adds: "There is little or no evidence that class size reduction by itself benefits pupils later in their school careers."
Also, he says, the benefits of a smaller pupil-adult ratio apply only when there is a small number of pupils for every qualified teacher. Increasing the number of non-teaching staff in the classroom does not have the same effect.
However, most studies on the effects of class size tend to measure outcomes purely in terms of literacy and numeracy.
By contrast, Blatchford argues, "small classes seem to promote more positive pupil attitudes, enthusiasm, confidence and ability to learn independently ... But these kinds of pupil outcomes have rarely been studied in research."
In particular, class size is likely to affect how much individual attention pupils get from their teachers. Several different studies found that the frequency of one-to-one contact between teachers and pupils - and, therefore, teachers' knowledge of individual pupils' needs - increased as class sizes shrank.
Other studies showed that the fewer pupils in a class there were, the better the relationships between teachers and pupils, and the more likely teachers were to offer pupils feedback on their work.
Research also indicated, albeit less strongly, that small classes led to easier classroom control, lower teacher stress and higher morale.
Such outcomes, however, are rarely measured in research studies because academic results are far easier to quantify.
This, Blatchford suggests, could explain the difference between research findings and teachers' professional experience. Teachers, he says, tend to insist that smaller class sizes will lead to more effective learning.
"My feeling is that this disparity is likely to have much to do with the fact that teachers have in mind a wide range of pupil attributes, covering academic achievement, but also pupil attitudes to learning and behaviour," he says, "while research on the effects of class size has been mostly directed at test scores of pupil attainment in mathematics and literacy."
Blatchford, P. "Class size: is small better?", Bad Education: debunking myths in education (Open University Press, 2012)
Peter Blatchford, Institute of Education, University of London. bit.lySQ5YiS
Smaller classes must be accompanied by appropriate teaching methods, otherwise the effect will be minimal, the University of London's Peter Blatchford says.
The Department for Education recently concluded that smaller class sizes have less effect on pupil achievement than alternative, less costly reforms.
But Blatchford points out that class-size reduction is often compared with initiatives such as one-to-one tutoring, peer mentoring and computer-assisted learning.
These, he says, offer distinctive methods of teaching, whereas reducing class size simply dictates how many pupils will be sitting in a room.
"A fair test ... would need to also take into account what teaching and instruction would be appropriate in classes of different sizes," he says.
There is a view, Blatchford concedes, that reducing class sizes automatically benefits teaching. However, evidence has shown that teachers do not always change their teaching style when faced with fewer pupils.
"There is a need for teachers to carefully consider ways in which they ... make the most of having fewer pupils," he says.
"It would be particularly valuable to concentrate on strategies for increased personalised, appropriate instruction ... but (also) to ensure that we do not see all the benefits of smaller classes in terms of increased opportunities for individualised teaching, rather than other pedagogical approaches."
For example, teachers should not neglect collaborative group work simply because they want to concentrate on individualised teaching.
Blatchford concludes: "The benefits of class-size reduction, should it be introduced, are not likely to be maximised without attention to effective teaching in small classes."