comes to establishing causal relations between class size, classroom processes and
achievement, and this is likely to be exacerbated by the use of a mixed method approach.
We have seen that one aim was to combine quantitative methods such as time estimates
with methods more able to capture the individual situation in classes and give expression
to teacher experience. We have in this paper been able to give scant expression to these
richer forms of data, and this reflects difficulties involved when seeking to include them
in analyses of causal relations. Clearly qualitative and more interpretative data have
purposes other than establishing causality, but if it is agreed that there is value in
integration then there is a need to think through more carefully ways in which this can
take place, and this would include ways in which relatively fine grained data on
classroom processes can be used in analyses of causal relations between a factor like
class size and pupils’ educational achievement and learning.
There are several limitations of this study. There may be other ‘outcomes’ and classroom
processes, related to class size differences, that were not covered. Any study has to be
selective and, although aspects of children’s attainment and classroom processes were
chosen carefully on the basis of previous research and our visits to schools, there may be
other features that are important. In the case of child ‘outcomes’, we deliberately
concentrated on the academic outcomes of literacy and maths, not the least because we
believe they are important indicators. We did cover areas of child development other
than attainment. We have, for example, seen that class size was related to student
inattentiveness and this can be seen as a basic requirement for learning. We also found
that work related interactions between children were affected by size of class. In the end
of year reports from teachers and the case studies there were a number of suggestions of
ways in which class size could be related to learning in a more general sense than
academic attainment. But there are no doubt effects of class size differences on other
aspects of children’s work, for example, in terms of more creative and artistic areas, and
in terms of other ‘outcomes’ like problem solving. Anecdotally we know that teachers
feel that time devoted to literacy, numeracy and science means that there is now less
time available for more creative and artistic activities, and this might be expected to be
especially true in larger classes. There is a case for looking more systematically at
dimensions we did not cover.
There is also the point that any kind of statistical analysis, no matter how sophisticated,
rests on quantification, and it may be that some essential processes and their effects
cannot be captured in this way. We have discussed this possibility elsewhere
(Blatchford, Goldstein and Mortimore, 1998). It may be that a smaller class size allows a
teacher to approach children in a more personalised and more humanistic way, but this
would be hard to measure and then enter into statistical analysis!
Another potential limitation of the research is the time when it took place. All
educational research is historically located and since it began there have been more
recent initiatives, including the rearrangement of the stages of education for young
pupils such that the ‘reception’ year (the year in school when children become 5
years) is now classified as the last year of the ‘Foundation Stage’. There is not space
here to enter into debate about this change but, whatever the merits or otherwise, the
question we do need to address is whether it affects the interpretation of the research
findings presented in this paper. It is difficult to answer this question with any
certainty, not the least because we do not have systematic information on just how
much change in reality there has been in the Reception year. However, the connection
24