I spoke out strongly (TES September 13) because the evidence then at my disposal suggested that the new Teacher Training Agency research was going to be uttering general statements from the study of individual cases. I am delighted to read in Anthea Millett’s article, and in the one by Claire Troft and Allen Andres, who are participating in the TTA teacher-researcher initiative, that I am wrong and that generalisation out of context will not happen. This suggests that a message that has been hammered out within the research community over the past 20 years about the essential difference between the study of a singularity and the search for generalisation is now widely accepted within the profession of teaching.

The key issue in the dissemination of a case study about classroom or school practice is whether other practitioners can relate the context of the case to their own situation. A case study cannot predict, but it may suggest. Relatability rather than generalisability is the methodological stance needed.

By contrast, of course, policy-makers do need general statements. A wiser government would, for example, have conducted substantial longitudinal studies of different kinds of nursery education on a scale and sampled basis which permitted general statements to be made, before introducing a nursery voucher scheme.

These matters are not academic pedantry. They arise from a view of educational research as being about judgments and decisions aimed at improving educational action. They recognise that there are two major audiences for such research: practitioners (for example, teachers, heads, inspectors) and policy-makers (for example, politicians, the Department for Education and Employment and local education authority officers). And overall they are concerned with the pursuit of truth.

Anthea Millett, and the teachers researching on the TTA scheme, should be assured that the higher education community of educational researchers welcomes the Agency’s commitment to teaching as a research-based profession. As the author of Creating Education through Research (1995) I hope I have made my personal commitment clear. Argument begins not with the aim, but on how it should be achieved.

Professor David Hargreaves, in his lecture to the Teacher Training Agency last year, asked “just how much research is there which ... demonstrates conclusively that if teachers change their practice from x to y there will be a significant and enduring improvement in teaching and learning?” This is a dangerous question because the answer “probably none” wrongfoots the researcher. The question should be: “how much research describes how individual teachers have tried out changes in their practice and improved the learning in their classrooms as a result?” The answer to this is “quite a lot, but much more is needed”. The action research movement, stemming from Stenhouse’s work on teachers as researchers (in An Introduction to Curriculum Research and Development, 1975) and developed by people like John Elliott, Jack Whitehead, Pamela Lomax, Marian Dadds, Bridget Somekh and many other teachers in higher education and in schools has shown how research methods linked to a commitment to improvement can enhance the quality of learning - in individual situations.

Professor Hargreaves comment was in the same vein as Lord Skidelsky, speaking a couple of years earlier in the House of Lords, saying “there is no theoretically based good practice which defines professional teaching.” This view fails to understand that teaching is not a technological skill, like word-processing, but is an art, like essay-writing. Every teacher has an extremely complex theoretical basis for his or her work - but it is personal.

That part of action research which is concerned with reflective practice engages teachers in exploring and articulating aspects of their practice. The result often leads the individual teacher to create for himselfherself worthwhile change.

This is why Jack Whitehead argues strongly for putting the “I” into research - which is, of course, anathema to the traditional scientist. I fear that Anthea Millett is dangerously near to the HargreavesSkidelsky position when she complains that “too much research stops short of working out the meaning of research findings for day-to-day practice”.

Research has demonstrated the significance of parents in relation to their children’s literacy development, for example, but it must be for the individual primary schoolteacher to decide whether and how to apply this in the context of his or her class.

Partnership between teachers and educational researchers is important in terms of the common aim of enhancing the quality of education.

While they explore the same territory they have different roles: teachers to educate children, researchers to pursue truth. Educating children is a formidable task but also, as the Scott Inquiry was told in a different context “truth is a very difficult concept”. Methodological issues, like the one discussed here, are a vital part of the pursuit of truth: to ignore them will damage the educational enterprise.

Michael Bassey is executive secretary of the British Educational Research Association