It's as easy as a, b or c

Is multiple choice key in an age of mass participation? asks Harriet Swain.

Philip O'Neill, head of English at Northumbria University, tots up the amount of work facing his staff as exams come to an end. "Each staff member would have up to 60 exam scripts to mark and the same number of 2,000-word essays. Then there are 10,000-word dissertations, and each staff member would have about ten of those. Most of us are external examiners, too, so we receive regular scripts for that. And then they have to be moderated - checked, double-checked and discussed." All this has to be done in about a week - possibly two.

With student numbers increasing, examiners' workloads aren't getting any lighter, nor are demands from students for careful assessment.

Student number 60 in the pile of exam scripts will probably have incurred a debt of thousands of pounds in taking a degree, and the mark awarded could determine how quickly it can be paid off - no matter that the examiner's main concern by then is how to keep awake. What is the answer?

Well, maybe lots of answers. Multiple-choice tests are used extensively in the US, particularly in admissions, and they are beginning to appear in Britain. Aptitude tests for admissions are becoming more common, and often rely on multiple choice as part of an overall emphasis on objectivity. Many science subjects use multiple choice, and the increase in e-learning also depends on this method.

A Department for Education and Skills consultation paper, Towards a Unified E-Learning Strategy , stresses the need for use of online assessment to improve efficiency, although it concedes that multiple choice is not suitable for all disciplines.

Multiple-choice assessment has advantages, particularly on courses such as business studies where there are often too many students for lecturers to assess regularly in other ways. This type of assessment is largely objective, and its efficiency means that students can be tested on a larger proportion of the syllabus than with traditionally set questions, making it easier to detect gaps in knowledge.

But there are disadvantages. First is the perception that multiple choice is less demanding than other forms of assessment. As Chris Rust, director of the Oxford Centre for Staff and Learning Development at Oxford Brookes University, puts it: "Students are quoted as referring to multiple choice as multiple guess."

There are ways of combating this problem - usually by using negative marking, whereby students are penalised for every answer they get wrong.

But a study by Laura Tatham and others at Manchester Metropolitan University into the use of multiple choice points out that it does tend to lead to higher scores. This is because it uses the full marking range, unlike traditional assessments, which tend to have a "glass ceiling" around the 80 per cent mark. The study suggests designing questions more carefully to ensure that the same level of achievement is needed as for a conventional exam.

Question design is another problem. Multiple-choice exams are easier to mark but more difficult to set. Rust warns of the dangers of questions unwittingly giving away the answer through grammatical inference and of questions answering those posed earlier in the paper.

He suggests that where multiple choice is most useful is in "formative assessment" - assessments that do not count towards a final degree mark but that can help students and teachers understand quickly where problems lie and discover how much is being taken in. He argues that the method of assessment is irrelevant. It is the quality of feedback that is important.

The key phrase, stresses the Quality Assurance Agency, responsible for ensuring high-quality assessment processes, is "fitness for purpose".

A spokesman for the University of Cambridge Local Examinations Syndicate says that while multiple choice is useful for testing small amounts of knowledge and, in its most complex forms, how people think, it is no good for testing more than one or two skills at the same time. "There is no reason why a degree, or part of a degree, should not be multiple choice, so long as the creators of the test are absolutely clear what they are trying to assess," he says.

Tatham, whose work referred specifically to law departments, says multiple choice is more useful for assessing higher-level learning than many people realise, particularly application and understanding of knowledge. "It is just a question of how you write them," she says.

Harvey Woolf, head of academic standards at Wolverhampton University, who is researching assessment in history, also says there is probably more scope for its use in the humanities and social sciences.

But as for how to help increasingly overloaded examiners, O'Neill says more money is the only answer.