The president of the British Psychological Society has lashed out at British university examinations, saying that results give a misleading picture of teaching quality and are often an inadequate measure of students' performance.
Stephen Newstead has called for nationally set exams in an attempt to harmonise major discrepancies between university exam results.
Professor Newstead, whose research into exams has looked in particular at student cheating, was addressing the annual meeting of the society in Brighton this week. In his presidential address he called for a radical overhaul of the student assessment system.
As well as his own research into the deficiencies of exams he has gathered research from around the country. He cited an example of two experienced markers differing by 70 per cent in the marks they awarded, and an essay being awarded an excellent first from one examiner and a borderline lower second/third from another.
Professor Newstead said that studies have shown inconsistencies between disciplines, gender bias and exam cheating. He called for more blind marking and more clear guidelines on student cheating.
He also called for marking schemes. "This is contentious," he said, "because they remove the discretion of the examiner. It would be quite threatening to people. I feel threatened myself."
National exams would ensure comparability between disciplines even if "only a part of the overall assessment" could be examined. "If we don't do this we will move to an American system where degrees are not standardised."
Professor Newstead, who works at Plymouth University, also criticised the use of students' exam results to measure their lecturers' performance. "It's a very indirect measure of teaching quality and may not measure it at all," he said. "Exams are a good measure only of certain aspects of student performance. What has changed is that we have started using exam results as a measure of teaching quality. That puts different pressures on them."
He said that lecturers are tempted to ensure that students do well. Institutions differ in their proportion of good results in ways that are "hard to explain in terms of intakes". In psychology, two apparently similar institutions produced 30 per cent and 70 per cent good degrees.
Meanwhile, his research shows that the number of firsts has gone up by 50 per cent since 1980, with most of the increase happening in the past eight years.
Professor Newstead said he was concerned that the external examiner system was overloaded and not functioning properly.