Four experiments, first used by Bigsby (1988 a), were presented to three groups of competent readers of varying reading ages and to nine developmentally dyslexic children. As in the initial study, competent reader data were analysed using group statistical analysis techniques together with some discussion of individual experimental effects and error levels. Competent reader data were also analysed, as were data from the dyslexic subjects, at a cognitive functional level and individual processing descriptions were obtained.
Results from the group analysis confirmed earlier findings that nominal matches engage an abstract letter identity (ALI) code which has a similar capacity to the visual code and that, when the ALI code of a reversible letter (b, d, p, q) is accessed, an extra processing step is probably involved. Individual processing descriptions revealed that several of the dyslexic subjects displayed major impairments in ALI coding while some of the competent readers showed mild inefficiencies. A developmental trend towards greater processing efficiency was noted, using both methods of analysis, for the competent reader sample.
It is concluded that results are consistent with those models of word recognition which depict ALI coding as a necessary prerequisite to word recognition.