| Abstract: | In this case study we investigated numeral transcoding in CM, a young man with developmental dyscalculia. Our initial assessment revealed that CM's difficulties in numeral transcoding occurred mainly on tasks requiring the production of Arabic numerals. His errors on these tasks were most often syntactically ill-formed strings in which the basic lexical elements are retained, but the overall structure is incorrect (e.g. nine hundred two thousand seventy → 92,70). We then implemented a training programme designed to overcome CM's syntactic impairment. Pre-training/post-training comparisons revealed significant improvement on the trained written-verbal-to-Arabic task as well as on an untrained spoken-verbal-to-Arabic task presumed to share the same Arabic numeral production process. A comparison of CM's numeral transcoding performance with control children showed that CM's performance improved from a third-grade level to approximately a fourthgrade level (i.e. the grade in which most children begin to display proficient number transcoding skills). Further analyses revealed that the syntactic errors produced by CM and the control subjects were inconsistent in terms of length, a finding that contrasts with the results of previous studies of numeral transcoding in children. In the general discussion we mention possible reasons for the discrepant outcomes across studies. |
