Average A-level grades of university entrants are up, but figures give little support to some myths about higher education, Bernard Kingston argues
National debates on educational standards regularly feature A levels, especially when increased passes at the higher grades are announced. The examination is accepted as a respected university entry qualification. It is used, too, as a measure of quality in The Times university ranking table, and it is now possible to measure the changing fortunes of this primary input over the period 1995-96 to 1999-2000.
But it is well known that an increasing number of would-be university entrants have other qualifications, including Level 3 vocational certificates and the international baccalaureate.
Vocational certificates figure strongly in the government's wider access policy, and more than half the intake at seven or eight of the post-1992 English universities have qualifications other than A levels. A number of these received significant extra Higher Education Funding Council for England funding for attracting students from disadvantaged areas - the so-called postcode premium. The Universities and Colleges Admissions Service tariff comes on stream this year as an attempt to show equivalence of entry standards. Once the tariff has been established and widely accepted in the sector, it may be adopted for The Times rankings, but for now A-level scores remain one of the nine quality measures used for UK universities.
A previous article on longitudinal trends ( THES , June 2001) examined the shift in degree class over the five-year period to 1998-99, and in particular the proportion of good honours (firsts, upper seconds and enhanced) awarded by each of the 97 institutions that appears in The Times Good University Guide. Overall, the rise was about six percentage points, probably less than many observers would have anticipated. At least with this primary output measure of individual success, the analysis was perhaps a riposte to those who preach that "more means worse". But within this modest rise, movements of 10 percentage points or more were detected for a group of 20 universities. In four cases (Glamorgan, Oxford, Hull and Luton), the rise was 15 percentage points or more.
For entry grades, the movements revealed are less dramatic, but there are still some intriguing differences in performance.
The data on which this analysis is based cover only the 84 English, Welsh and Northern Irish universities that have been in existence for the whole of the five-year period under review. The Scottish universities have been omitted because a number of them have very few A-level students. In The Times tables, we use as our working definition the actual A-level and AS entry scores (maximum 30 points where A=10, B=8 down to E=2) for first-year, first-degree students under 21 years of age.
For this cohort, the proportion of entrants with A levels at the 1999 intake was 76 per cent. The average A-level score for all universities in the analysis rose over the years 1995-96 to 1999-2000 by more than a full point, from 17.4 to 18.7. This equates to more than one grade of an AS or a shift towards a higher grade at A level. If a series of three-year rolling averages is calculated, to reduce any effects of selecting a particular base year, the increase is from 17.8 to 18.5. This increase is not surprising given the general increase in A-level performance by sixth-formers over this time.
Generally, the performance of individual universities has been fairly stable. Cambridge, Oxford, the London School of Economics and Imperial College, London, have headed the table in each of the five years. At the bottom, positions have been less stable, but North London and Thames Valley have been regulars in the lower four. One reason for the greater stability at the top is that the range is more spread out: the top 20 covers a range of about 8 points, compared with 4 at the bottom.
The pre and post-1992 universities form almost entirely distinct halves to the tables, with only Lampeter of the pre-1992 universities appearing among the post-1992 universities in the lower half in some years. Lampeter, Salford and Bangor come consistently at the lower end of the pre-1992 universities, while Bournemouth, Nottingham Trent, Sheffield Hallam and West of England are consistently near the top of the post-1992 tables.
The positions of individual universities are broadly consistent, but there are some differences in the changes over the period. The largest ten increases in the three-year rolling average are by 1.3 points or more, while at the other end of the spectrum some universities had a declining score. There are a number of possible explanations for this differential performance between universities.
- Are the best getting better? There is no relationship between the change in the average score and the score in 1995-96. Thus there has been no increase in divergence - the best are not getting better at the expense of the worst. This result may be affected by the fact that universities near the upper limit of 30 cannot increase by much, but only five universities have a five-year average above 26 points, so this is unlikely to be a major factor.
- Is age attractive? Entry grades at pre and post-1992 universities have increased by about the same amount, so the traditional universities have not been grabbing the best students at the expense of the newer ones.
- Does London cost too much? Entry grades for London and non-London universities have risen by about the same amount, so the expense and other aspects of study in London have not impaired the capital's ability to recruit students with good A levels.
- The price of expansion? Universities with larger increases in intakes over the period have not moved down the tables. Expansion has not been at the expense of A-level scores.
The only positive relationship identified is with the change in applications. There is a reasonably strong correlation (r=0.29, p<0.01) between the increase in applications to a university and a rise in the average A-level score. This is not a particularly surprising result - more applications allow universities to be choosier. Perhaps more surprising is the fact that the change in the number of applications per acceptance (a better measure of competitiveness) is rather more weakly correlated (r=0.21, p>0.05) with the change in average score.
So some of the more obvious possible explanations of the pattern of changes in A-level scores can be rejected. Unfortunately, data are not available to test hypotheses concerning expenditure on, or the quality of, marketing activity, but it may just be that some universities have been more effective at convincing potential students of their merits.
Rising A-level scores may have positive benefits for a university. There is a high correlation between the average A-level score and the proportion of students gaining a first or upper second-class degree (r=0.88 for the 1999-2000 data) and the Hefce measure of efficiency (r=0.81 for 1999-2000 data). Correlation does not necessarily imply causation, but it is plausible that high performance at A level by incoming students is likely to lead to a lower probability of dropping out and a higher eventual degree performance.
Bernard Kingston is a partner in Mayfield University Consultants, b.kingston@mayfield-uc.org.uk .
The Times Good University Guide 2003 , edited by John O'Leary, is now available from major bookshops or direct from The Times or HarperCollins websites. The author wishes to thank Nicola Bright of Bright Statistics for her invaluable help in preparing this article.
Universities' changing A-level entry scores, 1996-97 to 1999-2000
. . . .
1996-97 Top Ten
.
1996-97 Bottom Ten
. 1 Cambridge 29.6 1 North London 8.5 2 Oxford 28.7 2 South Bank 10.5 3 Imperial College .3 3 Thames Valley 11.1 4 LSE .3 4 East London 11.5 5 Nottingham 26.1 5 Sunderland 11.7 6 Bristol 25.6 6 London Guildhall 11.8 7 Warwick 25.1 7 Greenwich 11.8 8 Durham 24.8 8 Glamorgan 11.9 9 Sheffield 24.4 9 Teesside 11.9 10 Birmingham 24.1 10 Lincoln 12.0 . . . .
1997-98 Top Ten
.
1997-98 Bottom Ten
. 1 Cambridge 29.7 1 North London 8.7 2 Oxford 29.2 2 Thames Valley 9.3 3 LSE .7 3 Luton 10.0 4 Imperial College .5 4 South Bank 11.2 5 Bristol 26.4 5 Greenwich 11.5 6 Nottingham 25.9 6 Glamorgan 11.9 7 Warwick 25.9 7 East London 12.0 8 Durham 25.2 8 Sunderland 12.2 9 UCL 25.1 9 London Guildhall 12.3 10 Sheffield 25.0 10 Lincoln 12.3 . . . .
1998-99 Top Ten
.
1998-99 Bottom Ten
. 1 Cambridge 29.7 1 Thames Valley 10.6 2 Oxford 29.3 2 North London 11.1 3 LSE .9 3 Luton 11.2 4 Imperial College .8 4 London Guildhall 11.3 5 Bristol 26.7 5 East London 11.7 6 Nottingham 26.2 6 South Bank 11.8 7 Warwick 25.9 7 Brighton 12.0 8 Sheffield 25.8 8 Greenwich 12.1 9 York 25.4 9 Wolverhampton 12.2 10 Bath 25.3 10 Sunderland 12.2 . . . .
1999-2000 Top Ten
.
1999-2000 Bottom Ten
. 1 Cambridge 29.7 1 Thames Valley 10.3 2 Oxford 29.5 2 Luton 10.6 3 LSE 28.2 3 North London 11.3 4 Imperial College .9 4 Greenwich 11.5 5 Bristol 26.5 5 London Guildhall 11.8 6 Warwick 26.3 6 Teesside 12.2 7 Nottingham 26.2 7 Middlesex 12.3 8 UCL 25.8 8 Derby 12.4 9 York 25.5 9 East London 12.5 10 Bath 25.4 10 Staffordshire 12.7 . . . . . . .
Average A-level scores over five years to 1999-2000
Top 20
.
Bottom 20
. 1 Cambridge 29.5 1 North London 9.6 2 Oxford 29.1 2 Thames Valley 10.3 3 LSE .5 3 Luton 10.9 4 Imperial .4 4 South Bank 11.2 5 Bristol 26.1 5 Greenwich 11.7 6 Nottingham 26.0 6 East London 11.8 7 Warwick 25.3 7 London Guildhall 11.9 8 Durham 25.2 8 Sunderland 12.1 9 Sheffield 24.8 9 Teesside 12.4 10 UCL 24.6 10 Wolverhampton 12.5 11 York 24.5 11 Glamorgan 12.5 12 Bath 24.1 12 Lincoln 12.5 13 Birmingham 24.1 13 Brighton 12.6 14 Manchester 23.7 14 Middlesex 12.6 15 King's London 23.4 15 Derby 12.7 16 Leeds 23.4 16 Anglia 12.9 17 Queen's Belfast 23.1 17 Coventry 12.9 18 Cardiff 22.6 18 Staffordshire 13.0 19 Newcastle 22.5 19 Westminster 13.1 20 Exeter 22.4 20 De Montfort 13.2
Top Ten Post-1992
.
Bottom Ten Post-1992
. 1 Lampeter 13.7 1 Nottingham Trent 16.3 2 Salford 15.3 2 Sheffield Hallam 16.0 3 Bangor 15.8 3 Bournemouth 15.7 4 Bradford 17.3 4 West of England 15.6 5 Brunel 17.5 5 Northumbria 15.5 6 Essex 17.6 6 Leeds Metropolitan 15.5 7 Aberystwyth 18.1 7 Oxford Brookes 15.0 8 Goldsmith's London 18.4 8 Plymouth 14.7 9 Queen Mary 18.7 9 Liverpool John Moores 14.6 10 Swansea 19.0 10 Portsmouth 14.6
Most by the post
Hefce Recurrent Grant, 2002-03 Postcode Premium
. University Premium as a % of total teaching grant 1 Sunderland 2.7 2 Wolverhampton 2.6 3 Derby 2.3 4 Salford 2.3 5 Teesside 2.3 6 Liverpool John Moores 2.3 7 Northumbria 2.2 8 Staffordshire 2.2 9 Central Lancashire 2.2 10 Luton 2.2
Entry Grade 3-Year Rolling Average
. Data from 1995/96/
97/98 1996/97
/98/99 1997/98/
99/00 Change in 3-year average 1
Northumbria
14.5
16.4
16.4
1.9
2
North London
8.5
9.4
10.4
1.9
3
Bath
23.3
24.1
25.0
1.7
4
Bournemouth
14.4
15.3
16.0
1.7
5
West of England
14.9
15.7
16.5
1.6
6
Brunel
16.7
15.5
18.3
1.6
7
Sussex
20.3
21.0
21.8
1.5
8
Nottingham Trent
15.6
16.5
15.1
1.5
9
Portsmouth
13.9
14.6
15.3
1.4
10
Lincoln
11.8
12.9
13.1
1.4
11
Manchester Metropolitan
13.0
13.6
14.3
1.3
12
King's College London
22.8
23.6
24.1
1.3
13
Sheffield
24.2
25.1
25.4
1.2
14
Oxford Brookes
14.4
14.9
15.6
1.2
15
York
23.9
24.6
25.1
1.2
16
Warwick
24.8
25.6
26.0
1.2
15
Hertfordshire
12.5
12.8
13.7
1.2
18
City
19.5
20.7
20.6
1.2
19
University College London
24.2
25.2
25.4
1.2
20
UMIST
21.5
22.0
22.6
1.1
21
SOAS
20.6
20.7
21.7
1.1
22
Liverpool John Moores
14.1
14.7
15.1
1.0
23
Anglia
12.3
12.9
13.3
1.0
24
Loughborough
20.4
20.9
21.4
0.9
25
Kingston
12.9
13.5
13.9
0.9
26
Goldsmiths College
18.0
18.7
18.9
0.9
Newcastle
22.1
22.7
23.0
0.9
28
Royal Holloway
20.8
21.3
21.7
0.9
29
Brighton
12.5
13.0
13.4
0.9
30
Exeter
22.0
22.4
22.9
0.8
31
Bristol
25.7
26.2
26.5
0.8
32
Queen Mary
18.4
18.8
19.2
0.8
33
London School of Economics
.2
.6
28.0
0.8
34
Queens, Belfast
22.7
23.1
23.5
0.8
35
Surrey
19.1
19.5
19.8
0.7
36
Coventry
12.6
12.9
13.3
0.7
37
Westminster
12.8
13.3
13.5
0.7
38
Leicester
21.1
21.5
21.8
0.7
39
Huddersfield
12.9
13.5
13.6
0.7
40
Cardiff
22.3
22.8
22.9
0.7
41
Aberystwyth
15.7
15.9
18.4
0.7
42
Southampton
21.8
22.0
22.4
0.6
43
Glamorgan
12.1
12.3
12.7
0.6
44
Bangor
15.6
15.8
16.2
0.6
45
Sunderland
11.8
12.1
12.4
0.6
46
Essex
15.3
15.4
15.9
0.6
47
Leeds
23.1
23.4
23.7
0.6
48
Imperial College
.1
.5
.7
0.6
49
Leeds Metropolitan
15.2
15.7
15.8
0.5
50
Reading
20.0
20.2
20.5
0.5
51
East London
11.6
11.8
12.1
0.5
52
Swansea
18.6
18.7
19.2
0.5
53
Lancaster
21.4
21.9
21.9
0.5
54
South Bank
11.0
11.2
11.5
0.5
55
Oxford
28.9
29.1
29.4
0.5
56
Kent
19.6
19.9
20.0
0.5
57
Teesside
12.2
12.6
12.6
0.4
58
Central Lancashire
14.1
14.4
14.5
0.4
59
De Montfort
13.1
13.2
13.4
0.3
60
Cambridge
29.4
29.7
29.7
0.3
61
Plymouth
14.7
14.9
14.9
0.3
62
Manchester
23.6
23.9
23.8
0.3
63
Bradford
15.2
15.1
15.4
0.2
64
Nottingham
25.9
26.1
26.1
0.2
65
Middlesex
12.6
12.8
12.8
0.2
66
East Anglia
21.1
21.1
21.3
0.2
67
Birmingham
24.0
24.3
24.2
0.2
68
Central England
14.0
14.0
14.1
0.1
69
Wolverhampton
12.5
12.6
12.6
0.1
70
Greenwich
11.6
11.8
11.7
0.1
71
Sheffield Hallam
16.0
16.0
16.0
0.1
72
Durham
25.2
25.0
25.2
0.0
73
Thames Valley
10.1
10.4
10.1
0.0
74
Liverpool
21.5
21.8
21.5
0.0
75
Derby
12.7
12.7
12.7
0.0
76
Aston
21.5
21.3
21.4
-0.1
77
Staffordshire
13.1
12.9
13.0
-0.1
78
Salford
15.4
15.2
15.3
-0.2
79
Luton
10.8
11.1
10.6
-0.2
80
London Guildhall
12.1
11.8
11.8
-0.3
81
Hull
19.8
19.4
19.5
-0.3
82
Ulster
19.4
19.4
18.9
-0.5
83
Lampeter
13.9
13.4
13.4
-0.5
84
Keele
20.2
19.9
19.4
-0.8