Does ability to establish symbol-sound pairings mediate the RAN-reading relationship?
Mads Poulsen, Holger Juul & Carsten Elbro
Performance on tests to rapidly name letters and digits has been shown to correlate with reading. One possible reason is that these tests probe the ability to learn and automatise symbol-sound associations. However, most studies have not controlled for the amount of experience with the RAN-items, so it is unclear whether it is the experience or the ability to take advantage of the experience that is responsible for the correlation between RAN and reading. Paired associate learning tasks have been shown to differentiate dyslexics from controls, and to correlate with reading in unselected samples. We tested whether the ability to learn and/or established letter knowledge mediated the correlation between RAN and reading in a sample of preschool beginning readers.
105 preschoolers were administered traditional reading, letter knowledge, and RAN tests, and taught novel names for animals in a paired associate learning task. These animals were then used in a rapid naming task.
Preliminary results show that reading correlated with the amount of training required for learning the animal names (r=-.19, p=.06). RAN speed with the same animals did not correlate with reading, but the number of error produced during this task did (-.27, p<.01). None of the PA-measures mediated the relationship between a traditional RAN and reading (r=-.36), but controlling for letter knowledge did.
The results indicate that established letter knowledge to a larger degree than the ability to learn sound-symbol associations is responsible for the RAN-reading relationship in early reading development.