Jim Rose asks us for our views about various matters, such as how best to teach reading, but makes only one point in defence of the Office for Standards in Education study of reading in three inner London local education authorities (TES, November 8).
He points out that the OFSTED study judged pupil progress using test scores administered at one point in time, together with some background factors. The problem, as we explained in our critique, is that the absence of prior intake data for the same pupils precludes any judgments about progress or about the relative effectiveness of schools or teaching methods.
Education research has shown that valid conclusions about cause and effect require, at a minimum, longitudinal follow-up data on pupils. Jim Rose appears to believe that conclusions about the effects of particular kinds of teaching are obvious on the basis of a single day's inspection without the careful collection of long-term data.
It was precisely this shortcoming that caused us to prepare our original critique of the OFSTED study. We are certainly not trying to undermine what inspectors and others are doing to try to improve education. Nevertheless, if OFSTED wishes to carry out research, it needs to think more carefully about how this should be done.
It is in everyone's long-term interests that research conforms to recognised quality standards, especially when it may be used to inform policy. This implies taking expert advice prior to designing a study and subjecting research to peer group review, for example via seminars, before publication. Such is the norm and it is one way in which obvious weakness can be corrected. One of our aims in criticising the OFSTED reading study was the hope that, in future, OFSTED might raise the quality of its research by reforming its procedures.
PROFESSOR HARVEY GOLDSTEIN and PROFESSOR PETER MORTIMORE Institute of Education University of London 20 Bedford Way, London WC1