Select Committee on Education and Employment Minutes of Evidence



FURTHER MEMORANDUM FROM THE HIGHER EDUCATION FUNDING COUNCIL FOR ENGLAND (HE 146)

NON-CONTINUATION RATES AND SOCIAL CLASS

  1.  In our original submission we stated that, though we will only be sure when we have completed the full modelling exercise, it seems likely that some factors that appear to be associated with non-completion can, in fact, be explained by the age of students, their entry qualifications and subjects of study. To illustrate this we showed how the difference in non-continuation rates by Social Class were much reduced if we just selected young students with high A-level grades. In this note we provide further analysis which supports that point.

  2.  Table 3 below extends the tables 1 and 2 in the original submission.

Table 3

NON-CONTINUATION FOLLOWING YEAR OF ENTRY BY SOCIAL CLASS AND ENTRY QUALIFICATIONS (YOUNG FULL-TIME FIRST DEGREE ENTRANTS, 1997-98)
Social Class Per cent not in HE High
A-levels (23-30)
per cent
Mid
A-level (13-22)
per cent
Low
A-level (2-12)
per cent
Not
A-level
per cent
All entrants per cent
I25 10105
II36 11116
IIIn36 11137
IIIm36 11138
IV and V37 12139

  1.  The table shows that the differences in non-continuation rates between Social Classes is much reduced by separating them into different entry qualification groups. This approach will not completely take out all the effect of entry qualifications, because of the broad groupings used. This is particularly true of the "not A-level group" which collects together a wide range of qualifications. These differences are illustrated by table 4 which shows the average A-level points within each Social Class—entry qualification grouping. (Note the date used to generate table 4 uses the fine branded data used in the Pls. This should give a close approximation to the actual averages.)

Table 4

AVERAGE A-LEVEL POINTS BY SOCIAL CLASS AND A-LEVEL BAND (YOUNG FULL-TIME FIRST DEGREE ENTRANTS, 1997-98)
Social ClassHigh A-levels (23-30) Mid A-level (13-22)Low A-level (2-12)
I27.518.4 9.5
II27.118.2 9.3
IIIn26.918.1 9.2
IIIm26.617.9 9.3
IV and V26.717.9 9.1

  2.  The differences between the Social Class entry qualification profiles could, in theory, be reduced by taking narrower bands. However, the resulting small numbers would mean that random errors would confuse the picture. An alternative way of demonstrating how entry qualifications, with subjects, explain a large part of the observed differences in non-continuation rates, is to calculate "benchmarks" for each Social Class in the same way that the benchmarks are calculated for institutions in the performance indicator tables. Table 5 below shows the results of these calculations.

Table 5

NON-CONTINUATION FOLLOWING YEAR OF ENTRY BY SOCIAL CLASS WITH BENCHMARKS (YOUNG FULL-TIME FIRST DEGREE ENTRANTS, 1997-98)
Social ClassPer cent not in HE Actual per centBenchmark per centActual-Benchmark per cent
I5.15.9 --0.8
II6.36.8 --0.4
IIIn7.37.3 --0.1
IIIm8.08.1 --0.1
IV and V8.68.3 +0.2

  3.  The benchmark figures are what the whole population would have, had they had the entry qualification and subject profile of the particular Social Class. As can be seen, the actuals are close to the benchmarks, suggesting that most of the differences between Social Classes can be explained by these two factors. The difference between actual and benchmark increases for lower Social Class, and this could be due to the effect of, say, increasing financial pressure. (Note that the benchmark figures do not average out to the actual figures because of the effect of the students with unknown Social Class.)

  4.  However, it should be remembered that even the benchmark calculation involves a rather crude summary of the entry qualifications, so that this pattern of the residuals could be due to these limitations. The work we are currently undertaking involves getting a much fuller description of the entry qualifications, and applying sophisticated modelling techniques to allow for these and other factors. We will also be using a range of variables to act as proxies for the relative financial positions of students. As well as Social Class we will be using the relative affluence of their neighbourhood using geodemographic classifications, and on whether the students' parents are required to pay a fee.

Higher Education Funding Council for England

February 2001


 
previous page contents next page

House of Commons home page Parliament home page House of Lords home page search page enquiries index

© Parliamentary copyright 2001
Prepared 14 March 2001