Education: Testing: after the U-turn, which way next?: A radical rethink of the Government's plans for testing pupils was announced this week. But, as Judith Judd reports, an older type of test based on reasoning skills is gaining new popularity.

Click to follow
THE OFFICIAL hunt is about to start for a way of measuring pupils' progress through school, after ministers announced on Monday that they will investigate ways of making league tables fairer. But the unofficial search has been on for some time.

While controversy has continued over the Government's new national curriculum tests, some local authorities have been quietly setting their own tests and using them to draw up league tables. Unlike government tables, the local authority ones are 'value-added', taking account of the criticism that it is no use looking at how pupils are performing when they leave a school unless you also consider how they were doing when they came in. They compare GCSE results with scores in verbal and non-verbal reasoning tests taken between the ages of 10 and 13.

Many pupils took tests of this type in the Fifties and Sixties, but they fell out of favour in the Seventies and Eighties. Now, demand is booming. Reasoning tests are so popular that the National Foundation for Educational Research has just issued a new version. 'Sales have been phenomenal,' says a spokeswoman. 'They have increased by 50 per cent in the last year. The new tests have been administered to 120,000 pupils in the past three months.'

Different NFER verbal reasoning tests are set for pupils aged from seven to 14. One question for 10- and 11-year- olds asks them to underline the two words out of three that are closest in meaning (whisper, silent, hear). Another gives the code for a word and asks them to work out the code for another word (If TEAR is Q34M what is ATE?). Non-verbal reasoning tests ask pupils of the same age to decide how shapes are alike and to complete patterns of shapes.

More than 20 local education authorities use the NFER tests and others use similar tests. In the past, such tests have been used to measure pupils' progress. In the new world of performance indicators they are increasingly being used to measure schools.

Teachers are wary about league tabling based on results from these tests, but less uneasy than they are about the Government's tests. So could they provide the key to the improved league tables for which ministers are searching? They have clear advantages over the present national tests: they make allowances for pupils' ages, they are marked on a finely calibrated but simple scale, and are standardised. As David Jesson, senior lecturer at Sheffield University, who is working with Shropshire council on pupil progress using such tests, puts it: 'You know a pupil who scores 103 in Cornwall and one who gets 103 in London are performing at the same level.'

Research also shows that reasoning tests are good at predicting pupils' achievements in GCSEs. A confidential report for outer London boroughs completed recently by the NFER showed a close correlation between pupils' performance in verbal reasoning tests on entry to secondary school and their GCSE grades. A study in the Fifties produced similar results.

By comparing results at, say, 11 and 16, schools can be helped to improve the education of pupils. Shropshire uses an NFER test at 10-plus and plots pupils' GCSE performance against it. This is used with information about age and gender to show heads how different groups - lower-ability children, for instance - are performing. The information about how schools are doing is not published, though the council is discussing how it might be.

Keith Hedger, the county's senior adviser, says reasoning tests are a much fairer way of judging schools than the present government league tables, which only give raw results. The reasoning test gives information about prior attainment. 'This is the most precious bit of information you can have. If you have no prior attainment score you can't measure anything.'

Teachers, however, are not enthusiastic about the tests. Mr Hedger describes the Shropshire test as 'not brilliant'. Headteachers in the London borough of Merton, which measures progress in a similar way, are looking for new reasoning tests to replace the current ones. All 13-year- olds take two 10-minute tests 'very old-fashioned and reminiscent of the old 11-plus,' according to Chris Armstrong, a teacher at Rutlish school. Pupils are not told the results.

Tony Mooney, head of Rutlish, says the Merton tests have a poor correlation with GCSE and are biased against girls. His analysis of three years' results showed boys outperforming girls at the top of the scale.

His pupils are also critical. Robert Solheim, aged 14, who took the tests last year, says: 'I don't think they give a good view of someone's education. It's a very narrow view of intelligence.' He and his classmate Lee Baldwinson say there is not much motivation in doing tests when you never learn the results.

Mr Mooney, however, is convinced that the right type of reasoning tests must be used if reliable 'value-added' tables are to be produced. 'They are simple to administer and a good way of identifying under-achieving schools.'

Experts agree. Mr Jesson says that it would be better to have more detailed information than reasoning tests can supply. Even so, he argues, these tests allow us to see how young people make progress and help schools to raise standards.

Professor Harvey Goldstein, of London University's Institute of Education, says: 'Research in this area is in its infancy. I don't think using the government tests to compile league tables showing progress is on. No one has really thought through the problems of tracking each pupil through school. Verbal reasoning tests are the best we can do at the moment.'

Sir Ron Dearing, head of the Government's curriculum review, said on Monday that there was no point in looking for an ideal form of league table to show what schools contribute to pupils' performance. League tables based on reasoning tests are not ideal, but they offer a more reliable and acceptable way of measuring progress than any others yet devised.

(Photograph omitted)