Let's make it fair

Alternative methods of student assessment may work better than traditional exams

The exam season is here again and hard-pressed students and academics are once more asking if there aren't any better alternatives to traditional exams. Since modularisation, administrators have been frantically trying to timetable twice as many and those hiring out church halls for exams are contemplating doubled income.

At the same time, many of us working in assessment are questioning assumptions that exams are the fairest and most efficient methods of assessment. Madeleine Atkins, in a review commissioned by the Higher Education Quality Council of standards in 34 subject assessments, suggests: "Many lecturers were far from expert, accurate or reliable when assessing students' work."

Stephen Newstead, president of the British Psychological Society, has also been considering the reliability of exams. He is worried about inconsistencies between disciplines, gender bias and cheating, and cites serious problems of inter-tutor reliability with, for example, two experienced markers differing by 70 per cent for the same piece of work.

This tends to explode conventional wisdom that time-constrained unseen exams are ideal. Even when they achieve consistency and reliability of marks, they only assess a limited range of students' knowledge and ability. We can usually be sure that we are seeing the students' own work, but certainly not the best they can do. At their worst, traditional exams require stressed students using unreliable short-term memory to respond to often badly written questions under strict time constraints. What chance then to demonstrate logical argument, the application of theory to realistic contexts and a holistic approach to what they have learnt?

At the University of Northumbria, we are exploring the impact of innovative methods of assessment on students and staff, and we believe it is possible to use alternative forms of examination. Academics are often asked to prove that their new methods work well, yet there is little convincing evidence that the old ones are fit for purpose. The challenge is to demonstrate to traditionalists that new methods can be equally rigorous.

Lecturers throughout the country are developing different kinds of exams which include:

in-tray exercises where students receive a dossier of papers and a variety of tasks to work on in the exam room. They can work on sorting out what is really important from potential red herrings, and to cope with the unexpected in a way that simulates real practice.

open-book exams, where students can have access to texts or formulae sheets and then work on questions of interpretation or analysis. These reduce the reliance on rote-learning and test instead what students do with the information.

takeaway papers, where the questions set can be worked on at a more relaxed pace. Many students are used to using wordprocessors for coursework and find it difficult to write fast under exam conditions. Takeaway papers also permit access to reference sources and can produce more thoughtful and polished work.

case-studies, where the exam questions are based on scenarios or case materials provided before or during the exam, so they can apply knowledge in new ways to practical examples.

multiple choice questions or short answer questions, forming all or part of an exam paper, so that students can write briefly and demonstrate an understanding of a range of topics. Multiple-choice questions have a poor reputation, because people think they can only assess low-level skills and are open to guessing, but we only have to look at the sophisticated and taxing questions used by Open University courses to see that this need not be true.

Students are always open to new ideas. A computing student using open book exams says: "We can concentrate on what we're here for; learning things rather than simply memorising. I like the way you can concentrate on actually understanding the material instead of trying to memorise lots of facts ... the ideas are what you need to learn."

I'm not suggesting we abandon conventional exams altogether, but that we broaden the range, since any single assessment method disadvantages some students. We may have a long way to go, though, to convince the diehards that anything else can ever be as rigorous or as fair as they claim traditional exams to be.

I brought back a cartoon from New Zealand that sums up my argument. It shows an elephant, a penguin, a monkey a goldfish, a seal and a dog standing before a examiner who is saying, "For a fair selection, everybody has to take the same exam; please climb that tree!"

The writer is an educational development adviser at the University of Northumbria at Newcastle and co-chair of the Staff and Educational Development Association.