Computer marking splits unis

Almost half of university exams are now marked automatically by computers, it was revealed this week.

The high-tech revolution in grading students' work emerged from the first national survey of student assessment in higher education, covering every British university.

It means that many lecturers are escaping the burden of marking exam papers. And student learning is improved as a result, supporters of the trend claim.

But some academics are unconvinced of the benefits of "cyber- assessment", warning that it imposes "mechanistic and reductionist" values on marking techniques, and could erode standards.

Joanna Bull, senior research fellow at Luton University's Computer Assessment Centre, said the 750 survey replies highlighted both the benefits and limitations of computerised assessment, which has grown hand in hand with the use of increasingly sophisticated multiple choice-styled examinations.

Dr Bull, survey project manager of a consortium including Glasgow, Loughborough and Oxford Brookes universities, said: "Although the results are still at a very early stage, we think we can show that a number of factors are combining to make computer- assisted assessment increasingly important to universities."

Pressure on academic staff from rising student numbers was the most important motivation, but Dr Bull stressed that quality assurance procedures also favoured the technology.

"This is certainly a more objective form of marking as it is not subject to marker fatigue or other factors which can influence grades," she said.

Dr Bull said there was still a common misconception in universities that old-fashioned multiple choice questions were inferior to traditional essay-style exams. She said the technology had now moved on, and an expert system-style approach could provide very sophisticated problem-solving questions that could be marked by computers.

But Chris Grey of Cambridge University's Judge Institute of Management Studies, described the technology as a "sticking plaster" for under-funded expansion.

He said: "This imposes a mechanistic and reductionist set of values in marking which is at odds with the analytical qualities that employers want from graduates."

Liz McDowell, a senior lecturer at Northumbria University, said there was an almost emotional attachment to traditional exams in high-status universities.

"This is an area of dispute because many academics still want to separate learning and assessment, which is no longer reasonable," she said.

The survey shows a patchy picture across UK higher education, with some universities establishing institution-wide strategies to bring in the technology while others are deeply sceptical.

"There is a danger that some academics will feel they have lost control of the process of marking, which for many is a very personal activity. There can also be the misguided worry that jobs may be threatened," said Dr Bull.

Devising computer-assessed questions was a skilled activity requiring considerable training time, she added.

The benefits included quicker, more, and better feedback to students and to academics. This could be targeted to reveal problems such as common gaps in knowledge.

While new universities were often more open to the idea, there were many examples of computer marking in the older universities.

"It tends to be strongest in the science-based subjects, but now social sciences, geography, psychology and business studies are getting involved, and we expect more growth in the arts and humanities," Dr Bull said.