A specific serial order learning deficit as a causal explanation for dyslexia? Examining the role of general and within-task confounding variables

A specific serial order learning deficit as a causal explanation for dyslexia? Examining the role of general and within-task confounding variables

First Author: Eva Staels -- VUB
Additional authors/chairs: 
Wim Van den Broeck
Keywords: Dyslexia, Sequence learning, Attention, Reading Ability
Abstract / Summary: 

Objectives
Recently a specific deficit in STM serial order learning has been proposed as a new causal explanation for dyslexia. The researchers argue that because serial orderings of verbal and spatial elements occur in reading, a serial order deficit may explain reading problems in dyslexia. However, a number of methodological weaknesses can be noticed in these studies. First, they did not control for attentional functioning or IQ (general confounding factors). Secondly, they did not control for item STM when examining a specific serial order STM deficit. In our study we used a design in which item vs. serial order STM tasks and verbal vs. nonverbal item material are bifactorially manipulated (within task confounding variables). Furthermore, groups are matched using entropy balancing on a number of general confounding factors. The results of our study provide empirical evidence regarding the serial order STM deficit hypothesis as a causal explanation for dyslexia.
Methods
In a 2x2 within-subjects design 4 tasks were administered in 30 dyslexic and 30 control children: a verbal item STM task, a nonverbal item STM task, a verbal serial order STM task and a nonverbal serial order STM task. Both groups were matched on IQ and SES and attentional functioning was administered in all participants. Participants were attending 4th or 5th grade in primary school in Flanders. The results of the study were analysed using entropy balancing as a matching technique.
Results
The significant difference between the poor readers and the control group on the verbal serial order task disappears after closely matching both groups on IQ and attentional functioning. These results show that poor readers do not suffer from a specific serial order learning deficit. Differences between both groups found in previous studies appear to be the result of general confounding variables.