Automaticity, Retrieval Processes, and Reading: A Longitudinal Study in Average and Impaired Readers

TLDR

The study tracked word‑retrieval speed and its link to reading in 72 average and 11 severely impaired children from kindergarten to grade 2 using three reading tests and four continuous naming tasks with varied stimuli. Results showed that as retrieval automaticity increased in average readers, naming‑speed/reading correlations shifted from broad to highly specific, with the strongest links between graphological naming speed and lower‑level reading tasks, while impaired readers were consistently slower—especially on graphological symbols—and three dyslexic subgroups emerged, including a large globally impaired group and two smaller groups with early retrieval deficits and dissociated reading deficits.

Abstract

In this longitudinal investigation, the development of word-retrieval speed and its relationship to reading was studied in 72 average and 11 severely impaired readers in the kindergarten to grade 2 period (5-8 years). Subjects received a battery of 3 reading measures and 4 continuous naming tests with varied stimulus requirements. Results indicated that the relationship of retrieval speed to reading is a function of development and the correspondence between higher- and lower-level processes in the specific retrieval and reading measures. As automaticity in retrieval developed in average readers, naming-speed/reading relationships moved from strong, general predictions to highly differentiated ones. The strongest correlations were between naming speed for graphological stimuli and lower-level reading tasks. Impaired readers performed slower than average readers on all naming measures across all years, particularly on graphological symbols. 3 dyslexic subgroups emerged: the largest was globally impaired across all naming rate and reading tasks; 2 smaller subgroups had early specific, retrieval-rate deficits and dissociated reading deficits.