Treatment for phonological dyslexia targeting regularity effects

Abstract

Background: A treatment programme for an adult case of acquired phonological dyslexia is described. In addition to disproportionate difficulty in reading nonwords and predominately visual errors for real words, this case of phonological dyslexia demonstrated clear regularity effects (fewer irregular than regular words read aloud and regularised pronunciations for irregular words) on testing. Based on Hillis and Caramazza's summation hypothesis (1991, 1995), it was assumed that the participant's difficulty with irregular words arose from an impaired semantic system that produced equal activation of both the target and related lexical representations. With an impaired grapheme to phoneme conversion mechanism (as suggested by his inability to read nonwords), the phonological information associated with the target was insufficient to cause the correct lexical representation to reach threshold from among the alternatives. In this situation, the participant's phonological impairment formed the basis for his difficulty with reading irregular words aloud. The treatment programme for this participant therefore targeted GPC processes to strengthen the phonological information necessary for producing irregular words. Aims: The study aimed to determine whether the accuracy and promptness of oral reading for irregular words could be improved by targeting grapheme to phoneme conversion processes, and whether any effects would be observed on other standardised reading tasks. Methods & Procedures: The treatment programme focused on improving grapheme to phoneme conversion for irregular words controlled for frequency of occurrence and imageability through repeated exposures to multiple exemplars demonstrating target phonological contrasts. Written production and syllable parsing followed oral productions. A multiple-baseline across-behaviours single-subject design was used to evaluate the effects of the programme. Outcome & Results: Increased sensitivity to the production of two graphemes in low-frequency conditions was suggested by the participant's performance. Clear treatment effects could not be demonstrated, however, because of ceiling effects and unstable baselines. Improvements in reading comprehension, reading rate, and nonword reading were observed post-treatment. Conclusions: As the theory used to guide this participant's treatment also implicates increased thresholds for the semantic activation of target lexical items as a basis for difficulty in reading irregular words, future work should continue to assess the relative contributions of treating lexical semantic versus GPC processes in these participants. Studies should persist in attempting to circumvent the experimental problems that might arise during the early stages of recovery while continuing to exploit more recent theoretical accounts of reading to identify effective treatments.