Ethnic variation in higher education participation amongst males in the UK: the mediating effects of attitudes and prior attainment

ABSTRACT

This article reports on the findings from a longitudinal analysis of 2976 boys from the Next Steps (formerly LSYPE) data set. It unites the existing literature on ethnic gaps in attainment and higher education participation (HEP) to offer deeper, more holistic insight into the relationship between ethnicity and educational outcomes. The article offers a robust understanding of the extent to which ethnic variations in HEP are mediated by attitudes and attainment. Structural equation mediation models were used to investigate the link between ethnicity and outcome across a seven-year period. The analyses show specific mediated effects of attitude to school and attainment on ethnic variations in HEP for boys from certain BME groups relative to their White-British counterparts. The findings have implications for policy and practice, both in compulsory schooling and in higher education.

KEYWORDS:

• Ethnicity
• higher education participation
• attainment
• social class
• parental expectation
• structural equation model

Disclosure statement

No potential conflict of interest was reported by the author.

Notes

1 General Certificate of Secondary Education – a standardised exam taken at the end of compulsory schooling in the UK.

2 On the advice of the data owners (DfE 2011b), the weight used in the analysis was from the last wave (wave 7).

3 Significance testing using a Chi-square test is often one of the core ways of testing the appropriateness of SEM for the data. However, although the chi-square result is reported for transparency, this sort of testing was not suitable for this data set which is not multivariate normal and has a large sample size with accompanied degree of freedom. Such a combination usually leads to a significant chi-square result and the possibility of a type 1 error (when taken in conjunction with good results from the other fit indices as in this case) (Schermelleh-Engel, Moosbrugger, and Müller 2003).

4 Note that the presence of the asterisk denotes significance (p < .05) in this and all the tables included in this paper. Additionally, standardised coef are only included in this model for comparison with later mediation models.

5 Broken down between two pathways ATS → HEP and ATS → Attainment → HEP.

 

Additional information

Funding

This work was supported by the British Academy through their postdoctoral fellowship and the Quantitative Skills Acquisition grant.