Is a procedural learning deficit a causal risk factor for developmental language disorder or dyslexia? A meta-analytic review

Is a procedural learning deficit a causal risk factor for developmental language disorder or dyslexia? A meta-analytic review

First Author: Gillian West -- The University of Oxford
Additional authors/chairs: 
Charles Hulme; Monica Melby-Lervag
Keywords: Meta-analysis, Implicit Learning, Developmental dyslexia, Developmental Disabilities, Literacy development
Abstract / Summary: 

Impaired procedural learning has been suggested as a causal risk factor for developmental dyslexia (DD) and Developmental Language Disorder (Nicolson & Fawcett, 2007; 2011; Ullman, 2004; Ullman & Pierpont, 2005). However, results from the many studies examining this theory are highly inconsistent. We conducted meta-analyses on evidence from the five procedural learning tasks that have most commonly been used to test the theory: the serial reaction time, Hebb learning, artificial grammar and statistical learning, and the weather prediction tasks.

The review addresses the following questions:
1) Is there evidence for a deficit on procedural learning tasks in children with developmental language disorder or developmental dyslexia and are equivalent relationships found in correlational studies that examine samples representing the whole range of language or reading ability?
2) Is there variability in the severity of impairment in language disordered groups across studies and, if so, are moderators such as disorder type, severity of the disorder, modality of the task used, age or nonverbal IQ related to study results?

We found that studies using the serial reaction time and Hebb learning tasks yielded small deficits in language impaired compared with typically developing controls (g = -0.30 and -0.32 respectively). However, the serial reaction time task was not a reliable correlate of language abilities in unselected samples (r = 0.03). Larger group deficits were found in studies using artificial grammar and statistical learning tasks (g = -0.48) and the weather prediction task (g = -0.63). However, the severity of language and decoding problems explained negligible amounts of the variation in group differences in procedural learning between studies.

We conclude that current data offer no convincing support for the theory that a generalized procedural learning deficit is a causal risk factor for developmental dyslexia or developmental language disorder.