Children with dyslexia automatize temporal skills more slowly.

Developmental dyslexia is normally characterised by unexpected problems in learning to read for children of average or above average intelligence. In recent research, however, we have established that dyslexic children also show problems in the gross motor skill of balance1 , 2 and in speed of choice reaction to pure tones, even though their simple reaction speed is normal3. We interpreted the motor skill deficits as symptoms of incomplete automatisation of skill, in that the deficits were apparent only when the children were prevented from consciously monitoring their performance either through use of a dual task or by cutting out visual cues. The well-established reading and phonological deficits of dyslexic children are fully consistent with this 'Dyslexic Automatisation Deficit' (DAD) hypothesis, since phonological skills evolve through long exposure to, and use of, speech and of course reading requires the fluent interplay of a number of sub-skills4. Unfortunately, the evidence for the DAD hypothesis is indirect, and therefore we undertook a further study testing the long-term learning of a temporal skill. Since we had established the deficit in choice reaction but not in simple reaction speed, the natural experimental design was to investigate the time course of the blending of two separate reactions into a choice reaction. Initially, children were tested on two separate simple reaction tasks: a finger press to an auditory tone and a foot press to a visual flash. Then a choice reaction task was constructed, using both modalities of stimulus, and the child had to make a finger press to the tone or a foot press to the flash as appropriate. Naturally, initial performance on this choice reaction task is much slower than the simple reactions, but through extended practice one would expect the latencies to diminish to near those for simple reactions5. The critical prediction, therefore, for DAD was that dyslexic children would improve their performance more slowly than non-dyslexic children. Two groups of children participated: 11 dyslexic children around 15 years old and 11 non-dyslexic children matched for age and full IQ. All the dyslexic children had been diagnosed between the ages of 7 and 10, based on discrepancies of at least 18 months between actual and reading age. Their IQ levels fell in the normal to superior range on the Wechsler Intelligence Scale for Children and they had no known neurological deficit or primary emotional difficulty. All reaction time tasks were computer administered and …