ABSTRACT
For the reasons we gave in the last chapter, our review of psycholinguistic interventions in aphasia is largely restricted to single case studies in which the following conditions are met: the nature of an individual's disorder in terms of a psycholinguistic model is specified, the nature of the intervention used is itemized in detail, and a single-case experimental design has been used which allows some inferences to be made about the effects of the intervention. This is not to deny that therapy can achieve clinically significant results which are not necessarily reflected by statistically significant results, as Pring (1986) has pointed out (see also our comments in Chapter 8 on the preference in conversational analysis for qualitative rather than quantitative descriptions). If the cognitive neuropsychological approach to aphasia therapy is to achieve certain other aims, however, it is statistically valid results which are at present of most interest. Without statistical probability measured, the aims cannot be met of, first, convincing a healthily sceptical scientific community that aphasia therapy is effective, and, secondly, feeding back into psycholinguistic models information which is useful in their development.
