Modelling the impact of schools and neighbourhoods on the pupil progress and value-added from KS2 to KS4 A cluster analysis of secondary schools in England

This quantitative study adopts an innovative educational data mining technique, cluster analysis, to model the effects of secondary schools in England on their pupils’ attainment at KS4. The aim of this study is to explore how do state-funded secondary schools cluster on value-added attainment at the end of compulsory education, and how schools differ between clusters on school characteristics and pupil compositions. In addition, the sustainability of cluster membership across time is also investigated to examine how each cluster of schools facilitate their pupils on the progression from KS2 to KS4. Applying clustering methods to school-level administrative data, the School Performance Tables, and neighbourhood data from the 2011 Census in England, secondary pupils’ KS2- KS4 progress was examined in the context of schools and neighbourhoods. This study demonstrates the differences in KS2-KS4 progress between the five retrieved school clusters, to a large extent, are attributable to the intakes of secondary schools. a small number of higher- performing schools benefit from intakes of fewer pupils from disadvantaged socio-economic backgrounds, and more with higher prior attainment. By contrast, schools experiencing persistent low attainment at KS4 serve relatively more vulnerable pupils both at individual and neighbour levels. Such division of schools on KS2-KS4 progress sustains during the 2011/12 to 2014/15 academic years. Despite potential impacts of major policy changes, such as the Wolf review of vocational education in 2014, the size of school clusters remains similar over time. And the majority of schools stay in the same cluster, whilst some moved to adjacent performance clusters. In the 2015-2018 period, as the new headline measures and new GCSEs introduced, school clusters obtained from 2015 data are reviewed on Attainment 8 and Progress 8 measures. Consistent with previous research and findings from the pre-2015 period, some school characteristics and pupil composition factors play a key role in explaining the between- cluster difference in KS4 performance. Selective admission policy and school locations appear to be related to the pupil intake, whereas free school meal eligibility and first language provide a proxy to the socio-economic aspect of pupil compositions. It has been uncovered in this study that the ‘expected progress’ measures were unstable over time and inconsistent with Attainment 8 and Progress 8 measures that cover a wider range of subjects and account for prior attainment at KS2. The school clusters reveal a close association between KS4 performance as assessed by the EP and a school’s pupil intake. Applied on school-level data, cluster analysis can be used as a promising approach to investigate the reliability of school accountability measures, thus highlight the necessity of adjusting for contextual variations in the process of school effectiveness evaluation.

Name dc:type.qualificationname

Doctor of Philosophy (PhD)

Level dc:type.qualificationlevel

Doctoral

Grantor dc:publisher.institution

University of Cambridge

Year dc:date.issued

2021

Author dc:creator

• Wei, Yuxiu

Advisor dc:contributor.advisor

• Winterbottom, Mark

• Key Stage 4
• Cluster analylsis
• School Performance Tables
• Educational Effectiveness Research
• School Accountability

Licence

• https://www.rioxx.net/licenses/all-rights-reserved/

Language dc:language

eng

DOI dc:identifier.doi

https://doi.org/10.17863/CAM.87658

OAI identifier oai:identifier

oai:www.repository.cam.ac.uk:1810/340232