THE IMPACT OF SCHOOL SIZE ON PERFORMANCE

1. It is surprising that so little research has been undertaken into the relationship between school size, and pupil performance. The few studies which do exist tend to focus more on the pastoral welfare aspects of unit size, rather than on quantified pupil outcomes. By far the most useful publication is the July 2002 research from NFER, sponsored by the Local Government Association: "The impact of school size and single-sex education on performance". In its review of the literature, this document notes that "evidence about the impact of school size on performance in secondary schools appears to be virtually non-existent". It goes on, "we have not been able to find any English research on this subject". The only statistical evidence NFER could find came from the Netherlands/Sweden/USA (one study): the analysis did not reveal any statistical relationship between school size and achievement that was independent of student background characteristics.
2. One possible disadvantage of small schools is that they may not offer the same range of subjects as larger schools. The NFER looked at this, and reported that larger schools do offer a wider range of subject options in Science, and Design Technology subjects, and they were much more likely to offer multiple foreign languages (at GCSE level). Students in larger schools were move likely to take double rather than single balance Science.
3. In terms of performance, research undertaken by NFER suggested that the relationship between school size and GCSE outcomes proved to be curvilinear – in other words, after controlling for pupil, school and area background variables, it was pupils in medium sized schools (cohort of approx. 6-7 forms of entry) who obtained the best results. The optimum school size varied to some extent dependent on key variables: sex of pupil, prior attainment, and type of school (girls’, boys, or mixed: grammar or comprehensive). However, the report states that the general pattern was clear, namely that performance improved with school size up to a certain point, and then declined. Hence the best results were obtained in medium sized schools and the worst in small or large schools.
4. The report goes on to state that the reasons for this are not obvious, and although it is interesting to speculate, it is important to note that statistical analysis can only establish the nature of the link between size and performance, not the explanation for it. Association does not prove causality. For example, if large schools obtained good results, this may not be because they are large; it could be because schools which obtain good results become popular, and therefore grow in size.
5. On this basis, it is relatively easy to see why there might be a positive effect of school size. Moreover, smaller schools might offer a limited range of opportunities, and perhaps have limited resources. It is less easy to see why school size should have a negative impact after a particular size has been reached. One can only speculate that possibly larger structures become too complex, or pupils too easily "lost" within the system.
6. Whatever the reason, it is important to bear in mind the central finding of the NFER study, namely that the impact of school size, although statistically significant, is relatively small compared to the effect of other pupil and school level variables. In terms of average GCSE scores, the difference between schools of optimum size (see Para 4 above), and the very smallest or largest schools, was no more than 0.15 of a grade.
7. The NFER study looked at the effect of a Sixth Form on the rest of the school as a separate variable. A school with a large sixth form was associated with better than expected results, whilst a small sixth form appeared to have a negative effect. One possible explanation for this is that schools with large sixth forms are more likely to employ more specialist staff who might also teach Year 11 pupils. However, the causal relationship is unclear, and it is possible that in schools with good GCSE results, more pupils tend to enter the sixth form.

CONCLUSIONS

8. The tenor of the NFER report, as reported above, is that medium sized schools tend to perform better than the very large, or very small. However, the caveats must be stressed, particularly the marginal effect of the size variable.
9. It is difficult to draw any firm conclusions to support, or refute, the NFER central finding, from schools within Oxfordshire. Over recent years, the largest school in the county has been a consistently good performer (Lord Williams’s, Thame), as has one of the smallest schools (Langtree School, Woodcote). Interesting though analysis may be, it would be unwise to allow any tentative conclusions over the effect of school size on pupil performance to be a key factor in decision-making over the future configuration of King Alfred’s. If children are "lost" within the complex organisation or a large school, how much of this is, in reality, the product of large schools having unsatisfactory spatial and organisational structures, because of their inheritance?
   
   Return to TOP