This study investigated the ability of auditory and visual temporal processing measured before school entry (mean age 5.36 years) to predict early reading development in an unselected sample of children. There were 142 children at the first phase (Preschool), 125 at the second phase 6 - 8 months later (early Grade 1; mean age 5.94 years), and 105 at the third phase12 months later (Grade 2; mean age 6.94 years). There were similar numbers of males and females. Visual and auditory temporal order judgement (TOJ) and Temporal Dot accuracy (rapid visual sequencing task) measured at Preschool explained a significant percentage of the variance in letter identification (an important pre-reading skill) measured concurrently. These measures also predicted a significant percentage of the variance in letter and word identification (word reading accuracy) and reading rate (fluency) measured in early Grade 1, even after controlling for the effects of age, environment, memory, attentional vigilance, non-verbal ability, and speech and language problems. They also significantly discriminated between groups of children at Grade 1 who could and could not use phonological decoding to read non-words. By Grade 2, these Preschool measures accounted for significant variance in word reading accuracy and fluency and in non-word decoding. Only Preschool auditory temporal processing accounted for significant unique variance in the reading measures at Preschool or Grade 1, but by Grade 2, visual temporal processing (Temporal Dot) also accounted for significant unique variance. Temporal Dot accuracy also explained unique variance in the rate of growth in these reading measures across this period. These changes in predictive ability by the auditory and visual temporal processing measures were interpreted as reflecting developmental changes in their roles in reading as reading develops. Auditory temporal processing was important in early pre-reading and reading and remained important throughout. Visual temporal processing only became important in the later phase, possibly because of increasing need to analyse letter sequences. Preschool temporal and phonological processing measures accounted for approximately equal percentages of variance in the reading measures at Preschool and Grade 1, but by Grade 2, the Preschool phonological processing measures accounted for significantly more variance in all reading measures, except Pseudohomophone Choice (orthographic processing). Very little of the variance that was explained in the reading measures was common to temporal and phonological processing. The variance that each uniquely explained in reading was more important than the variance they explained in common. Therefore, utilising both temporal and phonological processing predictors optimised prediction of early reading skills. The study also showed there was significant linear development occurring in temporal processing from Preschool to Grade 2. The correlations of scores on the temporal measures from Preschool to Grade 1 were moderate. The relative position of children within the distribution on these skills showed moderate stability over the short-term, but less stability over the long-term. The majority of children who fell in the bottom quartile on the temporal and phonological processing measures at Preschool remained in the bottom half of the distribution on those measures by Grade 2. These children may represent those who are at most risk for reading difficulties. Letter Word Identification showed high stability from Preschool to Grade 2.
There was little difference in the percentage of variance explained in subsequent reading between temporal processing measures obtained at Preschool or Grade 1. However, performance on the Visual temporal order judgement task was more likely to account for significant unique variance in reading when measured after school entry than before. This was consistent with the expected developmental changes in reading. When measured after school-entry, phonological processing measures accounted for greater percentages of variance in the reading measures than when measured before. There were also developmental changes in which phonological processing measures were important predictors of reading skills. When measured at Grade 1, rhyme and alliteration detection and phonemic segmentation were the most important predictors. However, when measured at Grade 2, performance on the Rhyme and Alliteration task had reached ceiling, so would no longer be a useful predictor of later reading. These results were consistent with developmental models of reading and of phonological processing. The results provided support for a causal role of temporal processing in reading development. They also showed that measures of visual and auditory temporal processing obtained close to school-entry would be a useful addition to predicting risk of early reading difficulties. However, additional work is needed to determine the most suitable temporal processing measures for this younger age group.