Interview

Winds of change

Peter Tymms has often put himself in the eye of a storm as he contests official claims about education statistics and examination results. Yet, despite attracting national attention, he retains the sense of mission that began in his early years as a teacher.

The Centre for Evaluation and Monitoring (CEM) is a long way from London, yet the work of its director has sent shockwaves through government assessment policies. Last October, Professor Peter Tymms earned a place in the The Independent newspaper's list of 50 `national treasures' for producing "education statistics that invariably undermine and often contradict official claims". Whilst appreciating the accolade, acting as whistle-blower has not always been easy, and changing the direction of key policies has often proved a slow process. He also has to be wary of treading on toes which could cost valuable contracts. "There is academic freedom but you would not like to see the university in disrepute," he explains. CEM is on the periphery of Durham University campus and is the largest educational research establishment in a UK university with an annual turnover of around £5.5m. Founded 25 years ago, it provides institutions, education authorities and government agencies with monitoring systems and on-screen tests as well as research and evaluation. "We now have a staff of 70 and work with thousands of schools, processing data from a million children a year," says Tymms. His career began as a science teacher at a state boarding school in Zambia, 200 miles from the nearest Tarmac road. "My ambition was to try to teach people about science, to get people to understand more about the world in which they live." The job provided a sharp contrast to his own public school education on the edge of the Yorkshire Dales, which was followed T by Cambridge University where he read natural sciences and captained the university gymnastics club. After three years in Africa, he took a teaching job in a secondary modern in Easington, County Durham. "I taught science there in what many would regard as a fairly tough, working-class mining community." Perhaps Tymms' calling to work among poor communities stemmed from his upbringing as the son of a Church of England rector. At independence, seven years before he began teaching there, Zambia had just four graduates and there was a strong desire for education.

Inspiration and direction

A Masters degree in education at the University of Newcastle completed in 1986 provided the springboard for his academic career. "Carol Fitz-Gibbon was running parts of that course and that proved to be very important." Professor Fitz-Gibbon had initiated the A-Level Information System (ALIS) in 1983, which provides performance indicators for post-16 students across all sectors of education. After completing his Master's degree, Tymms spent a year on secondment at Newcastle as a research associate with ALIS. In 1992, he became a permanent lecturer and started the Performance Indicators in Primary Schools (PIPS) project. Although the idea originated from his links with ALIS, he had no background in primary school teaching. "It grew out of a realisation that there was no professional monitoring system for primary schools." He believes his background as a secondary school teacher was an advantage because he was "not carrying the baggage that everyone else does". Between 1995 and 2000, dramatic improvements in Key Stage (KS) 2 results were recorded nationally but assessment undertaken for PIPS showed no improvements at all. "We found other people had (recorded) the same failings as well. I wrote about this and got chapters in a book about it but nobody was taking any notice." Opinions started to shift after his findings were published in the British Educational Research Journal in 2004. "I had looked at all the data we had, local authority data and studies other people had done; there were many studies saying the same thing, that the results were showing exaggerated rises. Basically, people were in denial. People were feeling very good that these test results were rising." Although he acknowledges that the Qualifications and Curriculums Authority has still not changed its official position on KS2 as a result of his findings, he says they are no longer challenged by people such as Sir Michael Barber, a key education advisor to Tony Blair when Prime Minister. Tymms does not blame teachers for the inflated attainment levels at KS2. "Standards were being set incorrectly ¬ that was the root of the problem. The solution is to anchor the test so that it does not change from year to year", he says. Although this might seem an open invitation for teachers to prepare pupils for the test, he says this has been done successfully in other parts of the world.

Damaging over-optimism

Hurt feelings were an inevitable consequence of showing there were no actual improvements after all but Tymms felt over-optimism could distort the future direction of education policy if allowed to continue. "My view was that we were living a lie and that the chickens would come home to roost eventually." The debacle over KS3 test marking by ETS last summer, which subsequently led to the test being abolished, has convinced him that the KS2 test is doomed as well, even though Children's Secretary Ed Balls insists that they are still needed for accountability of primary schools. Tymms says: "I think the writing's on the wall. Key Stage 1 went some time ago and this year we've seen Key Stage 3 go. Key Stage 2 is the only one remaining; it can't go on for much longer. We never knew the Berlin Wall was coming down until it did." Apart from widespread opposition to the KS2 test from within the teaching profession, he argues that it has had long enough to confirm its worth. "It has not proved to be the lever to raise levels that Barber and others thought it would be. The disaster of ETS running the tests just keeps in people's minds that nobody can do this ¬ it's too big a beast to run."

How PIPS works

Tymms founded the Performance Indicators in Primary Schools (PIPS) in 1993, before Key Stage national testing had been introduced. Variations in pupil progress quickly emerged between the 12 volunteer schools. "The key was feeding the information back to the schools; it was the kind of information that schools can't get in isolation," he says. Tymms argues that the key difference between data produced by projects such as PIPS is to do with accountability. "PIPS is for teachers and for schools. You are doing assessment in order to find a problem. It means you can spot the kid who has got a reading problem when five years old and not wait until they are 12 years old." With national testing, because the schools are accountable for the results, he says there is a tendency is to try to hide such problems.

Addressing concerns

The initial focus of PIPS was on mathematics among year six pupils, but PIPS now tracks a number of aspects of primary schooling. The development of peer learning in Fife, Scotland, is a current example of this. Peer learning is based on the belief that it benefits both the children doing the teaching, because existing learning is reinforced, and the children being taught because of the one-to-one attention they receive. "We've used assessments to show that peer teaching is effective," explains Tymms. PIPS is translated into numerous different languages and is used by a third of Scottish education authorities. "There are other tests around but there is nothing else operating in quite the same way," says Tymms. "Other countries are surprised at what we're doing but one of the things that works in our favour is that schools have control over their own budget. In many other countries it is actually controlled by some central body."

Future trends

Tymms believes the next major change to testing worldwide will be a move from paper to computers. Although such tests take longer to generate, he argues that a computer adaptive test (CAT) is superior because it is easier to modify since syllabus changes. Results can be fed back more quickly and questions can be selected to match individual capabilities. "You can engineer CAT to make sure that each pupil will come out with 60 per cent or higher," he explains. CEM is spearheading developments in this area with its Interactive Computerised Assessment System. This is designed for five to 11 year-olds but Tymms says interest in the approach is even coming from examination authorities. He also believes that assessment will increasingly move into less cognitively-based areas. In April, CEM's Internet Testing Unit began piloting a project to assess the mental health of 40,000 year four and eight pupils following research suggesting one in 10 children suffer mental health problems.

Under pressure

Tymms supports the NAHT's campaign to abolish the Key Stage 2 tests, arguing that their link with league tables and tough inspections have had unfortunate consequences. He argues the pressure to achieve level four, or above, results in the weighting of resources in favour of years five and six. "A year-six teacher does not have as large an impact in the long-term as a reception teacher. But they get children to get through statutory tests and that is important. PIPS has actually demonstrated this; we are literally able to compare progress from one year to another." He believes another consequence of the existing system is that, even with value-added weighting, league tables are prone to distortion, particularly when the levels are dependent on teacher judgment, as at the end of KS1 and now KS3. Because schools are ranked by the levels, teachers are encouraged to give higher levels where possible. Tymms says this leads to "some odd results which will have consequences for the value-added scores at the end of the next Key Stage". But he is cautious about scrapping Key Stage 3 testing. "The advantage of the decision is a decrease in high-stakes testing. The fear is that pupil failure will go unnoticed and schools' judgements will be based on opinion."

 
Tymms is dismissive of the suggestion that single-level tests, where teachers put pupils in for tests at a particular level when they are felt to have reached that level, are a viable alternative. "Most people I know in the psychometric world don't think this will work in England. Teachers are quite good at making judgements but they do make mistakes. If a pupil is put in for a level four and the teacher's got it wrong, it may end up meaning more testing than before." He says additional problems would be created by having to interpret data from tests which pupils were taking at different times. When he started PIPS among year six pupils, there was no Key Stage national testing. "One of the first teachers I dealt with said she did not believe they would sit for half an hour doing a test. Now it has gone too far the other way." However, he does not believe that the schools can simply depend on the type of research carried out by CEM, partly because of pressure from parents to know how their children are performing. He also thinks teachers benefit from having two separate sources of data to show how their class is progressing.

New skills needed

Tymms favours some form of objective, independent assessment but believes that, as a nation, we lack the necessary skills in psychometrics to provide this. He is optimistic that the new Masters degree in assessment, being rolled out in six universities, which leads to Chartered Educational Assessor accreditation, will help plug this gap. "The Government needs to know the standards of education across the country. I think it can do that with a sampling system; instead of assessing 600,000 pupils, you can do it with 5,000." Tymms is critical of the way special needs are identified, arguing that there is a clear bias towards middle class children. "Dyslexia is very common to find in independent schools and, bizarrely, less common in very tough areas. I think it's because special needs is related partly to how much push you get from parents." He also argues that it is wrong to place the same value on all A-level subjects when science, mathematics and languages are harder than those in the arts and humanities. "I think it is a great shame that we have allowed standards to drift. I am a great believer in A-levels; they spark a life-long interest in subjects. Physics and chemistry really came alive for me at A-level." But, he says, A-level currency has clearly been devalued when institutions such as the law school at Durham University now use a different test to select students. He does not, however, feel that creating a common level for each subject is workable, arguing that the solution is to clarify the relevant difficulty of each subject to the people who make use of A-levels. "It already happens in two or three places in the world. They have not always worked perfectly, but I think admissions' tutors would be able to follow this way of selecting." The prolific output of published works from Tymms shows no sign of abating. One of the latest books he jointly edited is on techniques for monitoring the comparability of examination standards. Although he believes people would be reassured by how effective these techniques are, he acknowledges there is a limit to what can be achieved. Despite his remarkable success since embarking on an academic career, much of what originally motivated him to become a science teacher still remains. "I like developing tests, analysing data and teaching," he says. "I miss teaching and get withdrawal symptoms when not in front of a group of people."

CV Age: 59 Qualifications: BA in Natural Sciences, Cambridge University; MA in Education, University of Newcastle upon Tyne; PhD Christ College, Cambridge University Employment includes: 1971-75 Science teacher Nchelenge Secondary School, Zambia 1975-1984 Chemistry teacher, Easington Comprehensive 1985-1990 Lecturer in Chemistry, Peterlee College 1990-92 Lecturer in performance indicators and evaluation, Moray House Institute of Education 1992-95 Lecturer in Education, University of Newcastle upon Tyne 1996-99 Reader in Education, Durham University 1999 Professor of Education, Durham University 2003 Director of CEM, Durham University