The purpose of this study was threefold: (a) to identify a combination of predictive measures that correlate with reading achievement, (b) to examine the predictive accuracy of these measures, and (c) to determine the most accurate time frame for test administration in kindergarten. One hundred and three kindergarten students from three schools participated over a period of two years. Measures representing letter identification, phonological awareness, phonological memory, and rapid automatized naming were administered in the fall and winter of the kindergarten year. Reading achievement was measured at the end of grade 1 using measures that included passage comprehension, fluency, sight-word recognition, and phonemic decoding. Five predictive models representing a combination of the predictive constructs were analyzed. The model combining letter identification, phonological awareness, and rapid automatized naming was identified as the best predictor of early reading achievement. There was no practical, significant difference between the fall and winter testing time frames. These findings hold important implications for predictive research by clarifying the importance of administering standardized measures that reflect the reading process. Most important, the results can provide practitioners with information for identifying the children most in need of early reading interventions.
Read full abstract