Measuring the determinants of quality in UK higher education: a multinomial logit approach
| DOI | https://doi.org/10.1108/09684880110399059 |
| Date | 01 September 2001 |
| Pages | 121-126 |
| Published date | 01 September 2001 |
| Author | Timothy Rodgers,Deb Ghosh |
| Subject Matter | Education |
Measuring the
determinants of quality
in UK higher education:
a multinomial logit
approach
Timothy Rodgers and
Deb Ghosh
1. Introduction
The numbers of students in UK higher
education increased dramatically from the
early 1980s onwards. For example, full-time
equivalent numbers increased by about
77 per cent from 1980 to 1992 (Mallier and
Rodgers, 1996). Constraints on public
expenditure at the time also led to demands
for greater cost efficiency, as was outlined in
the 1987 and 1991 Education White Papers.
This emphasis on student numbers and costs
can be seen as having changed the incentive
structures faced by the UK's higher education
institutions, a change which made the need
for accurate and objective measures of quality
all the more apparent.
From 1980 to 1992, the average cost of
teaching a student fell about 13 per cent
(Mallier and Rodgers, 1996). However, it
must be recognised that efficiency
improvements are not the only potential
reasons for the fall in average cost; as
institutions could also adjust the mix between
research and teaching, and/or reducing the
resourcing of, and therefore potentially the
quality of, teaching. The significance of this
latter point is reinforced by the findings of
Ghosh and Rodgers (1999), who show that
teaching quality is expensive and, therefore, in
theory at least, an institution could increase
cost efficiency by reducing educational
quality.
It is clear that the new incentive structure
faced by higher education institutions meant
that the quality of teaching needed to be
protected. However, if it is to be protected it
needs to be measured accurately, and if we
are going to compare the teaching output
quality of different institutions, it has to be
recognised that on grounds of equity, it is
necessary to take account of differences in the
quality of the student intake that different
institutions recruit. A ``level playing field'' will
only be achieved between institutions if we
can measure the relative importance of quality
significant variables over which the institution
has control (e.g. teaching hours), and quality
significant variables over which it has no
control (e.g. the academic potential of
students). This measurement problem is the
The authors
Timothy Rodgers and Deb Ghosh are both at Coventry
Business School, Coventry University, Coventry, UK.
Keywords
Higher education, Students, Performance,
United Kingdom
Abstract
A large number of correlation analysis studies of UK data
have suggested that there is a weak relationship between
student input quality (for example, ``A''level performance)
and degree performance. This paper re-examines this
relationship using a different statistical method, namely, a
multinomial logit analysis. Unlike the earlier studies, we
find strong explanatory variables for degree performance.
For example, it is found that an increase of one in average
``A'' points, other things being equal, increases the
probability of a first or upper second-class grade (a ``good
degree'') by 5.8 per cent, while student dissatisfaction
with a course will reduce this probability by 15 per cent.
Potential students should also note that taking subjects
like geography and history, other things being equal,
increases the probability of a ``good degree'', while
taking, for example, law, reduces it.
Electronic access
The research register for this journal is available at
http://www.mcbup.com/research_registers
The current issue and full text archive of this journal is
available at http://www.emerald-library.com/ft
An earlier version of this paper was presented at the
5th ICIT International Conference, Singapore,
Easter 2000. We acknowledge some useful
comments received at this presentation. The usual
disclaimer applies.
121
Quality Assurance in Education
Volume 9 .Number 3 .2001 .pp. 121±126
#MCB University Press .ISSN 0968-4883
To continue reading
Request your trial