This is because, thirdly, a diagnosis of an innate defect in the brain may cause learned helplessness in student and tutor alike (Kerr 1999, 2001a, 2001b). And this is because such a diagnosis is a maladaptive attribution – an attribution which disempowers and disadvantages. The literacy difficulty is, by its own definition, being attributed to an innate deficit within the student, which cannot be ‘cured’ and which can barely be overcome. Extraordinary measures and Herculean efforts will be necessary if it is to be, at any rate. The condition, it has logically to be presumed, will forever make the learning of literacy much slower, less certain, more improbable and more difficult than it would be for a ‘normal’ person. To accept this is to descend into the damp and fetid cellars of educational pessimism where learned helplessness grows like a fungus, consuming pleasure, motivation, enthusiasm, confidence, curiosity, engagement, stamina, expectation and performance. This fungus will, of course, infect everyone - family, teacher and taught alike.
In this section we will consider the main research-based scientific theories concerning the origins and manifestations of dyslexia. First, and fundamental, we consider sample selection.
The most basic aspect of research into any particular group’s characteristics is the selection of the sample to be studied. In order to study a particular group of people (for example diabetics, vicars, redheads or management consultants) the characteristics by which membership of the group is awarded must be defined as closely as possible. Criteria would, ideally, include all group members and exclude all non-members. If the study were to be of management consultants, for example, it would not be sufficient simply to include everyone who came to work on a particular train dressed in a dark suit. Such a sample might include several management consultants, but it might also contain dentists, advertising executives, I.T. consultants, white goods salesmen, up-market pickpockets, domestic science teachers, junior ministers and many others. Results of research on such a sample would not be dependably valid in respect of management consultants. It would not have been carried out solely among people known to be management consultants. We would not dependably reach a true sample of management consultants using this blunt selection process. Our research results will not be defensible unless we apply better, tighter sample selection criteria.
Thus it is, of course, with research into ‘dyslexics’. Before any meaningful examination of ‘dyslexics’ is carried out, the criteria for selection must be made clear. Who is ‘dyslexic’ and who is ‘normal’?
Assessment for dyslexia for the purposes of scientific research almost universally continues to regard the intelligence/achievement discrepancy criterion as the single pathognomic indicator of dyslexia and researchers either use it themselves, or rely for diagnosis on educational psychologists who use it. (A pathognomic sign is one which, even occurring alone, is sufficient absolutely to indicate a particular syndrome.) The discrepancy criterion is elucidated as follows: The IQ of a person is measured, and then their performance at some literacy skill or skills - reading, for example. The decision is made to consider someone as ‘dyslexic’ where there is a discrepancy of more than such and such an amount between the performance expected from a person with their particular measured IQ, and their actual performance in administered tests. The discrepancy is often expressed as a reading age discrepancy (a discrepancy between their expected reading age according to IQ and their reading age measured by actual reading performance).
Considering a particular discrepancy between measured IQ and performance on norm-referenced literacy tasks to be pathognomic for ‘dyslexia’ is problematic on two counts. The first is the increasingly frequent finding that a population defined as dyslexic by an IQ/achievement discrepancy criterion does not, in the event, differ reliably or importantly from the general population (Fletcher 2003, Fletcher 2004, Fletcher, Denton & Francis 2005, Fletcher et al 2005, Miles & Miles 1999, Samuelsson et al 1999, Siegel & Himel 1998, Stanovich & Stanovich 1997, Stanovich 2000, Stanovich 2005, Stuebing et al 2002). The discrepancy criterion is, to put it plainly, no longer held to indicate ‘dyslexia’ or any other neurological learning difficulty reliably, or even at all.