Education: Agonising over who'll make the grade: Donald MacLeod sits in with assessors who determine how many candidates will pass or fail their exams

'The purpose of this meeting is to attempt to maintain comparable standards for A-level awards,' announced Les Bolam, an Associated Examining Board examiner.

His words opened a session that would decide the fates of around 750 candidates who sat the board's sociology A-level examination in November. Their papers had been marked. Mr Bolam and his colleagues were now deciding what mark would be acceptable as a pass, and where the boundaries would be for different grades. The same procedure is followed for examinations in all subjects at GCSE or A-level.

These grading meetings are at the centre of the controversy over whether exam standards are falling or schools are performing better. When, last September, inspectors reported there may have been a 'gradual erosion of standards', it was this grading process that they criticised, not the actual question papers or lax marking.

John Patten, Secretary of State for Education, called on the exam boards to review their procedures; the boards dismissed the inspectorate's evidence as 'flimsy'.

The question as to whether standards are sliding or not hangs on whether boards are allowing weaker candidates to gain higher grades, or passing candidates one year who might have been failed in a previous year. The grading meeting decides where those lines should be drawn.

At the AEB's sociology A-level meeting, Mr Bolam worked alongside four other members of the awarding committee, all of whom had already marked some papers from last November's batch. They were advised by chief examiners for each of the two papers, with reference at each stage to statistics from previous papers.

Around the table in the board's headquarters at Guildford, Surrey, Karen Houlihan outlined the board's guidelines on grading, but made it clear that the decisions rested with the committee. They would be called upon to justify them if they stepped outside the guidelines.

Last November's entry was slightly down on the previous year's, although the number of A- level sociology candidates has grown from 18,000 to 26,000 in the past couple of years. Nearly three- quarters of entrants are female. The pass mark shifts slightly each year because no two exam papers are the same.

As the committee turned to Paper 1, Tony Lawson, the chief examiner, said he felt it was comparable to last year's, and that no single question attracted all good or bad answers. He recommended a minimum pass mark of 29 out of 75. Ms Houlihan said that in the summer the pass mark had been 27, but to pass the same percentage of candidates this time would mean a pass mark of 30. This difference caused considerable agonising.

The five started reading scripts that straddled the proposed boundary - marks between 24 and 32. They had scripts from the June 1992 candidates who received the same range of marks. Comparisons were not straightforward, however; a candidate who produced two good essays and an appalling one might score the same as one with three mediocre answers, so it was difficult to find a typical grade E or grade A essay.

The silence was broken by occasional comments and chuckles. Mr Bolam complained about the number of split infinitives. 'I wish they would teach them the difference between 'amount' and 'number' - it's getting on my nerves,' said Steve Walker, another examiner.

As they moved into the A and B area the mood lightened. 'This lot are really refreshing to read,' said Lyn Ashley. 'This is an interesting one on religion,' put in John Worgan. 'Not this one,' moaned Mr Walker. 'It's accurate but scores 'N' on interest.'

There was general agreement that candidates scoring 30 were definitely 'in'. Of the 29s, Mr Worgan felt that one was in and one was not. Ms Houlihan worried them by pointing out that this year the average mark went up, indicating, she said, that candidates found the paper slightly easier.

They wanted to stick to 29 however, so Mr Lawson was asked to justify recommending that mark. 'That was just my gut reaction to the marking,' he replied. The committee debated whether the paper was easier, whether this batch of candidates was better, whether teachers now have more experience of the new syllabus in its second year. Opinion hardened around a pass mark of 29. 'I don't think there has been any slippage of standards,' said Mr Walker.

The same process was repeated for the B/C and A/B boundaries before they moved on to Paper 2. Only half a dozen candidates made the top grade. Most were re- sitting, and the entry was weaker overall than in the summer.

Bill Sugrew, chief examiner for the second paper, said marks were up compared with November 1991 because students had demonstrated the skills of interpretation and evaluation which sociology A- level demands in addition to knowledge. That reflected changes for the better in the classroom, he argued.

Mr Lawson maintained: 'Slipping standards implies a massive conspiracy along the line, which simply cannot happen. In our case, rising pass rates mean an enormous candidature and a lot of teachers have been putting in a lot of work developing these skills, and it is beginning to pay off.'

One factor in rising pass rates may be the growth of sociology as a 'respectable' A-level in schools. This has happened in response to student demand: Thatcher's children are, it seems, eager to learn about society. The chief examiners believe it is partly an unintended result of the national curriculum - if pupils are given no choices until 16, they opt for something new then. The only A-level subject growing faster than sociology is psychology.

By mid-afternoon the committee's decisions meant that 48.6 per cent of candidates had passed, 4.6 per cent had reached the B boundary, and 0.8 per cent (six candidates) had received A grades.

At that stage, Ms Houlihan pointed out that 45.3 per cent had passed in the November 1991 exam, and the increase went over the board's guideline figure of how many candidates should pass compared with the previous year. The examiners, however, stood by their judgements. As Mr Worgan said, it was all over one mark. So Mr Bolam drafted a brief statement to the board arguing that the evaluation skills displayed in the answers they had read were better than in the previous year. Candidates had genuinely performed better.

Nevertheless, more than half had failed. A lot of young people had emerged from a two-year course, and probably another five months studying to resit the exam, with nothing to show for it. Standards, however, had been maintained.