Do educational researchers provide useful insights for teachers? Chris Woodhead, chief inspector for schools, clearly believes that they do not. He wants “to rescue the debate about education from the convolutions of academic discourse” (TES, February 26).

In deference to Mr Woodhead, I shall reply to his comments in plain English. In 1979, I gave up English teaching to become a contract researcher at the National Foundation for Educational Research. Many contract researchers are jacks-of-all-trades. During my 20 years, I have worked on projects as diverse as maths, reading, gender, citizenship and counselling.

Now, I will admit that the books that have had the deepest impact on my own views of education have not been research reports. Imaginative literature has been much more important. Plato, Rousseau, Dewey and Stenhouse are my four points of reference; Tom Brown’s Schooldays, Hard Times and Kes provide my contexts; Dibs and Pygmalion underpin my theories of learning; Just William, Tyke Tiler and Boy keep my sense of childhood alive; and Miss Trunchball’s maxim “If it’s fun, it isn’t learning”, makes me laugh.

So, if literature is more important to me than research reports, am I writing myself out of a job? I don’t think so.

Research reports play their part. But they are not instruction manuals for teachers. They merely contribute to the climate of debate in which policy decisions are made.

Research is not a fixed concept, it changes over time. In the past 30 years it has gradually been democratised so that now some of it is directly or indirectly in the hands of schools.

My first research job was a quantitative study - and I had to wade through stacks of computer print-outs. Such research provides broad-brush pictures of things that may influence a child’s education - parental qualifications, housing, income, mother tongue, school type, teaching approach.

The findings were of no immediate interest to teachers. But for policy-makers (and school inspectors), this kind of data is the sine qua non within which their judgments should be embedded. No case study, which is effectively what Office for Standards in Education inspectors produce, is worth a candle without an understanding of how the educational landscape fits together.

In the mid-Eighties, I worked on qualitative research. This was an exciting departure for those researchers whose hearts remained in the classroom. I saw myself as an anthropologist - making systematic observations, keeping meticulous field notes, discussing observations with pupils and teacher, gradually building a dense and textured picture of the complex worlds within a classroom.

I felt that research was where it belonged - in the classroom. But I quickly saw the flaws.

The major flaw was the disruption and stress I caused as I blundered through the delicate web of classroom interactions with my new video equipment. I soon realised that what I was observing was very different from the everyday lesson.

The imperfections of the observational method of research should be of real concern to Mr Woodhead. His inspectors impose much greater strains on their “subject” than researchers who enter classrooms in a genuine spirit of inquiry.

Now, I work alongside teachers in an action research project. The emphasis is on practical problem-solving through action and reflection, or if Mr Woodhead prefers it, through “tinkering” and thinking. It may be true, as he asserts, that all teachers tinker, but tinkering without thinking is of no value.

“Tinkering” is most effective when it is done in teams drawn from different disciplines - rocket science requires both engineers and physicists. However, Mr Woodhead is right to fear that some teachers in this kind of practitioner-research could be seduced into “the debilitating belief that there are experts out there in the universities and town halls who know more than good teachers know”. But mostly they won’t be.

The biggest danger comes from the Research Assessment Exercise which determines an institution’s level of funding. It puts pressure on universities to produce more publications - it is much quicker and cheaper to use teachers as data-collectors rather than to involve them in the research as participants.

Where research is school-based and in the collective grasp of teachers, governors, parents and heads, it is an extension of the democratic process. In this research, the school identifies a problem, looks for relevant theories and research, identifies a range of strategies, tries out the most likely and monitors the effects.

But they can’t do it alone. For one thing, they don’t have the time. As a parent-governor of a secondary school, I have been invited to contribute to thinking about disengaged boys. For an hour every few weeks, after school has finished, a small group of teachers and myself focus on how to foster a better learning climate for these boys.

My role is to provide up-to-date information about what is going on elsewhere. I do this freely and willingly - not least because I have a 10-year-old son who will soon be joining the school.

This kind of research deserves the support of universities - and of Mr Woodhead. Indeed, it may well be one direction for research, as university education departments begin to see themselves as more directly in the service of schools.

Research is not an opposition between practitioner and theorist but a partnership between those trying to make things work in school and those with time to access a different knowledge base.

Such partnerships, of which there are many examples here in Cambridge, seek genuine parity of esteem between all participants and are a healthy development in the relationship between universities and schools.

Jenny Gubb is a contract researcher attached to the University of Cambridge School of Education