Training phonological reading in deep alexia: does it improve reading words with low imageability?

Abstract

The purpose of this study was to examine whether training phonological reading using a bigraph‐syllable pairing method (e.g. pa ‐ /pæ/) instead of grapheme‐phoneme pairing (e.g. p ‐ /p/, a ‐ /æ/) resulted in improved reading of words with low imageability and paragraphs in an individual with deep alexia. In the treatment, Friedman and Lott's bigraph‐syllable correspondence (e.g. fa ‐ /fæ/) training procedure was adapted and used. The results indicated that the treatment effect generalized to the reading of words with low imageability, which was not reported in most previous studies. Oral reading accuracy and comprehension accuracy of paragraphs also improved. It is suggested that bigraph‐syllable correspondence training is effective in improving low imageability word reading, possibly because it can provide more substantial phonemic cues necessary to read words with low imageability (which have limited semantic information associated with them) than the grapheme‐phoneme correspondence training method.

• Deep alexia
• phonological reading
• imageability
• bigraph‐syllable pair

Notes

1. In fact, ten CV bigraphs and ten VC bigraphs were composed of three grapheme strings because of the use of consonant clusters (e.g. ull, ick, cha, shi,). In addition, nine CV bigraphs and 17 VC bigraphs could constitute an actual word (e.g. do, in, at, she), the pronunciation of 15 of them matching the target syllable production (i.e. whereas do was trained with the alternative pronunciation /d/, at was trained with the matching pronunciation /æt/ in treatment). The distribution of the bigraphs constituting an actual word was comparable across the sets. There were nine such bigraphs in Set 1, nine in Set 2, and eight in Set 3, with each set containing five bigraphs with pronunciation matching that of the target syllable production.

2. On rare occasions, more than one consonant group was introduced in a single session. This occurred when PT found the first consonant group to be unusually easy to learn and requested more items to learn, and when the second consonant group to be introduced consisted of only two items.

3. Note that PT's pre‐treatment performance on the PALPA Word Rhyme Judgment subtest suggested generally intact segmentation skills.