Traditional exam results are not necessarily the best way to measure student potential. Harriet Swain reports on efforts to improve admissions procedures that also help to widen access for non-traditional applicants.

Next week, thousands of young people will supposedly reach a turning point in their lives. Newspaper front pages will display the tears and whoops of joy that greet A-level results in what is the most high-profile period of the academic year. But increasingly these results are only part of the story.

Pressure on universities to recruit students from a wider range of backgrounds, combined with a growing number of applicants with top A-level or Scottish Higher scores and impressive CVs has led most institutions to look beyond traditional recruitment methods.

Assessment now takes place over months and involves candidates' school and personal backgrounds, enthusiasm for their subject and suitability for ÖйúAƬ. Higher education minister Margaret Hodge caused a furore when she suggested earlier this year that A levels were not the only way to measure potential, but within the sector her comments were not new.

ÖйúAƬ

Extra money from the funding councils for institutions taking students from disadvantaged backgrounds has been one reason for institutions to consider ways of assessing potential rather than past performance. Another is the worry that by concentrating too much on school exam results, universities may be missing out on the best brains. Research by Jeff Odell at Bristol University found that students from low-achieving schools subsequently performed better at degree level than those with the same grades from high-achieving schools.

Jane Minto, director of the Oxford colleges admissions service, says: "It is always the case that A levels or any school-leaving qualifications are primarily indicative of the ability to do well in public examinations."

ÖйúAƬ

This sparks another concern - that the increasing number of young people attending ÖйúAƬ, many of whom are getting heavily into debt for the privilege, should be those best equipped to benefit from three years of concentrated study, rather than simply being those who are good at exams. Research taking place at Manchester University is designed to find new ways of assessing applicants, in part by identifying what makes a successful student. Existing students will shortly take part in tests, interviews, focus groups and questionnaires to find out how their attitudes, personality, preferred learning styles and personal circumstances affect their ability to do well. The university is also trialling different types of psychological tests with local schoolchildren to see how useful they are in measuring potential.

Judi Turner, who is running the three-year project, says: "Within a lot of the faculties there is a realisation not only that some young people whose educational experiences are compromised may benefit from ÖйúAƬ but also that some with excellent attainment at further education are not suitable for ÖйúAƬ."

A further incentive for universities to change recruiting methods comes from looking beyond university to employment. Students expect to think about what job they are going to do once they leave. This is particularly true in vocational subjects and has sparked most interest in medicine, where a number of UK schools are piloting a variety of psychometric tests for use in admissions. John Hamilton, in charge of medical curriculum development at the University of Durham, Stockton Campus - which has been trialling tests delivered over the internet - says: "It is fairly clear now that ordinary academic qualifications are not a satisfactory assessment of people's quality as doctors."

What this means for admissions officers is paying more attention to personal information and teachers' references on application forms. For many, the first step is identifying those students who need further investigation, perhaps because they are at a poorly performing school, are mature applicants, or have achieved their results against the odds. These may then be invited to interview or to take further tests. Because it takes more time than simply ticking off the number of top GCSE grades and sporting captaincies, many institutions choose to focus on particular groups, such as local applicants.

At Nottingham University, for example, admissions officers use school league tables to help them assess applicants from schools within commuting distance of the university. They pay particular attention to those from the worst-performing schools, looking closely at their personal statements and references and at their achievements compared with those of their peers, also bearing in mind that some disadvantaged pupils can be found at relatively good state schools.

Penelope Griffin, widening participation officer at Nottingham and a coordinator of the Russell Group's widening participation association, says: "If you are an admissions tutor and have 3,000 applications on your desk you are going to be brisk. We are diverting off a subset and saying that on these we will take extra care."

It is up to individual schools of the university how they use this information. Some will invite certain students in for interview or ask for a piece of written work. Others will make the same offer to all students but bump those identified as disadvantaged up the queue, others will make a slightly lower offer. Of Nottingham's 32 schools, 30 use some kind of "alternative" admissions method.

Bristol University is going a step further. This year, for the first time, it has set targets in each subject for the number of students recruited from state schools, areas with a traditionally low uptake of ÖйúAƬ, mature students and lower social groups. These targets are dictated by the benchmarks set by the ÖйúAƬ Funding Council for England, which in 2000 set Bristol's benchmark for the proportion of state-school students at 73 per cent, compared with its actual figure of 55 per cent. Bristol hopes to be at least three-quarters of the way to the higher figure within the next two years. Faculties recruiting targeted students are rewarded with extra cash.

ÖйúAƬ

In the history department, which was singled out for praise by Hodge, standard offers have been abandoned completely. Offers can now range from two Es to three As and enormous importance is attached to the teacher's reference. In addition, all local and all mature students are invited for interview. Pat Rayfield, director for widening participation at Bristol, says one result of looking beyond traditional admissions criteria is that several departments have reverted to interviewing more candidates. "We are looking at how to identify our target students more closely," she says. "But overall it means we are having to look more closely at all students until we have those systems in place."

Admissions officers generally are placing greater importance on personal contact with candidates. Bristol is one of a number of institutions that host summer schools and will pay special attention to candidates who have attended them, especially as most attendees have been identified by their teachers as high achievers against the odds. Many new universities also run compact schemes, where they work with students over several months - helping to develop a portfolio of achievements, arranging visits to the campus, providing one-off lectures - and guarantee a place to all participants who achieve the necessary exam grades.

The universities of York, Sheffield and Leeds run a version of this through the "White Rose" Compact Scheme, which offers special consideration to applicants recommended by teachers, including providing special liaison officers and reviewing their cases "sympathetically" if they narrowly fail to make the grade.

ÖйúAƬ

While some get to know applicants, others believe objective testing is the answer to identifying candidates, although institutions admit to exploring the idea much more readily than to using it. Certain departments at Nottingham set tests for selected candidates and at Bristol the physics department has been using aptitude tests for two years in some cases. This autumn the plan is for at least one department in every faculty to get all its students to sit the tests as a way of tracking their effectiveness. Use of aptitude testing is also being considered at Queen's University Belfast. Oxford is exploring the possibility of having joint tests in some subject areas and is watching developments, but Minto says: "There is a sense in the system that candidates are nowadays doing a lot of exams and we don't necessarily want to add to that."

Research by the Sutton Trust on aptitude tests, which suggested they were no better than A levels in assessing pupil's potential and were most likely to benefit white males, although they did measure "something different", has left the subject unresolved.

Oxford is actively considering testing students applying to study medicine - the one area in which extensive trialling of tests is already taking place elsewhere. The tests draw on research carried out in Australia, but taking the final step to using them in this country will be a big one because of institutions' fears that top candidates may choose to stick with the traditional routes. "Someone has to bite the bullet and say they are going to use it as a definite tool," John Hamilton says. For him, recruiting candidates from non-traditional backgrounds is essential for the good of all student doctors. Having someone who lives in the kind of area of social disadvantage that many patients come from can help an entire class engage with what that means much more quickly, he says. "There is much more to widening access than politicians recognise."

Testing times for computer science

For the second consecutive year, the computer science department at Cambridge University is piloting a 90-minute multiple-choice paper to test "thinking skills" in admissions. The paper, which includes questions to test numeracy, verbal reasoning, logic and ability to prioritise information, was developed in response to difficulties in deciding between applicants all predicted top grades.

Robin Walker, chairman of Cambridge's directors of studies for computer science committee, says there was evidence that school exam performance did not correlate with subsequent aptitude for computer science.

Some colleges are using the tests to help with assessment, while others are not yet allowing the results to influence their decisions.

The University of Cambridge Local Examinations Syndicate, which designed the test, is considering its use in other subjects since none of the questions is specifically related to computer science.

Last year, the results of the 290 students who took the test, all with three grade-A predictions, varied from 18 to 48 out of 50, but it is too early to say whether the results correlate better with future university achievement.

Oz test to find good doctors

David Powis, associate dean of teaching and learning in the faculty of health at the University of Newcastle, Australia, has been developing tests with medical schools in England and Scotland designed to build up a personality profile of candidates.

The tests contain questions measuring personality, ethics and cognitive skills. For example, candidates will be given a dilemma, such as their wife is dying and the only way they can pay for the expensive medication needed to save her is by stealing. The test will examine how candidates resolve the dilemma.

More than 10,000 candidates have been tested so far in Australia and Britain.

Powis says he believes the test could soon be used but a long-term study is still needed to find out how useful it is in producing better doctors.

Peter Lampl argues that we should move from a predicted A-level to a post-qualification system with complimentary aptitude test. Do you agree with him? to join in the debate in The THES Common room .

ÖйúAƬ