This article reports results from 1- and 2-year follow-up tests for children with reading problems who had previously received 25 hr of training in 1 of 2 computer-based remediation programs. Both programs included accurate speech and decoding support during story reading, but different supplementary training, either in explicit phonological processes or in comprehension strategies. Children initially trained in phonological skills made greater gains in phonological awareness, phonological decoding, and untimed word recognition at the end of training. The other children, who had spent more time reading stories with speech feedback for difficult words, gained more in time-limited word recognition. One year later with no further training, the explicit phonological group remained stronger in phonological decoding and phoneme awareness, but its level of word recognition was no greater than in the other group. There were no significant group differences after 2 years. Both of the computer-based training programs had yielded rapid growth in word recognition standard scores during 50 half-hour training sessions over 4 months, although much more training is needed to bring many children with reading disability into the normal range.

Two hundred second- to fifth-grade students (aged approximately 7 to 11 years) spent 29 h in a computer-assisted remedial reading program that compared benefits from accurate, speech-supported reading in context, with and without explicit phonological training. Children in the "accurate-reading-in-context" condition spent 22 individualized computer hours reading stories and 7 small-group hours learning comprehension strategies. Children in the "phonological-analysis" condition learned phonological strategies in 7 small-group hours, and divided their computer time between phonological exercises and story reading. Phonologically trained children gained more in phonological skills and untimed word reading; children with more contextual reading gained more in time-limited word reading. Lower level readers gained more, and benefited more from phonological training, than higher level readers. In follow-up testing, most children maintained or improved their levels, but not their rates, of training gains. Phonologically trained children scored higher on phonological decoding, but children in both conditions scored equivalently on word reading.

One hundred twenty-two second- to fifth-grade (7- to 11-year-old) children with reading difficulties studied phonological awareness with or without explicit attention to articulation and with or without manipulation of sounds. They all studied identical phonics and read stories on the computer with speech and decoding support for difficult words. Regular-instruction controls received regularly scheduled language-arts or reading activities. After 40 h of training, children in all three trained conditions outperformed controls on all tests except math. Conditions that manipulated sounds showed advantages over the condition without explicit practice manipulating sounds, but only on the two measures of phonological awareness. Articulatory awareness training yielded no unique benefits during this training period. Individual differences in response to treatment related to initial levels of phonological awareness, naming speed, IQ, and grade. The similar outcomes of the three conditions suggest that specific variations in good phonological training may be less important than once thought for most children with reading difficulties.

Phonological training with an articulatory component (n=25) was compared to phonological awareness training without a specific speech-motor component (n=20) for second- to fifth-grade children with reading disabilities. No significant differences were found between the two kinds of training on the various outcome measures, except that the nonarticulatory group was slightly superior on phoneme deletion and spelling production, suggesting some small-scale overall advantages for the training with more manipulation of sounds. An interaction on nonsense word reading may suggest that the lowest children on phonological awareness benefited more from the articulatory training, whereas reading disabled children with relatively higher initial skills in phonological awareness may have benefited more from the nonarticulatory training. However, further research with a larger sample is needed before firm conclusions can be drawn.