Suppression of individual-item naming variance in the RAN-reading relationship

Presented at the 21th Annual Conference of the Society for the Scientific Study of Reading; Santa Fe, New Mexico, July 16–19 2014.

Athanassios Protopapas,1 Angeliki Altani,2 & George Georgiou2
1 University of Athens, Greece
2 University of Alberta, Canada

Purpose: To understand the concurrent relationship between discrete word naming and word reading fluency. Word fluency can be decomposed into a lexical access component, assessed with individually presented (discrete) word naming, and a serial execution component, assessed with simultaneously presented (serial) digit and object naming (RAN).
Method: 107 children in Grade 2 and 107 in Grade 6, native speakers of Greek from the general student population, were administered serial and discrete (a) word naming, (b) digit naming, and (c) object naming; plus a serial alternating naming task (RAS digits-objects).
Results: Serial word reading was the dependent variable in a series of regression models, starting with serial RAN and discrete word predictors. In both grades, inclusion of discrete naming tasks with the model predictors produced negative coefficients and caused the coefficient of discrete words to increase, consistent with suppression. The difference was statistically significant in Grade 2 only, in which it also caused an increase in the coefficients of serial RAN. A model with only RAS and serial RAN predictors was not significantly worse than the model including discrete words.
Conclusions: Phonological access is not the crucial factor underlying the RAN-reading relationship, because phonological access is also required in discrete digit and object naming but is effectively taken out of the equation through suppression. Orthographic processing is not crucial because individual word naming can be effectively substituted by RAS. These results highlight the importance of scheduling and controlling the sequential processing of successive words in a list.