Georgiou-Martinez-2021

Interestingly, the effect size for the CA-DYS comparison in reading comprehension in our meta-analysis (g = 1.43) is almost double the one reported by Reis et al. (2020; d = 0.72). This is likely due to the fact that our meta-analysis included dyslexia studies with younger participants. We take this finding to mean that, in adulthood, some of the individuals with dyslexia have likely developed mechanisms to compensate for their poor word reading skills when completing reading comprehension tasks (e.g., Birch & Chase, 2004; Deacon et al., 2012; Parrila et al., 2007; Pedersen et al., 2016)

However, we also found that individuals with dyslexia were performing significantly worse than their RL-matched controls (g = 0.64; a moderate effect size). This is interesting because the reading-level matched design (Bradley & Bryant, 1978; Bryant & Goswami, 1986) lies on the assumption that the dyslexia group and a group of younger children have a similar reading level. If younger children are matched to older dyslexic individuals on their reading ability, we would expect them to be matched on all reading tasks (including reading comprehension) and not just on one or two reading tasks. A similar issue was recently reported by Parrila et al. (2020a) in a meta-analysis with studies from consistent orthographies. This finding suggests that when researchers say they matched their samples on one reading task, we cannot assume that they matched them on all reading outcomes.

Our moderator analyses revealed a significant effect of orthographic consistency (differences being larger in languages with low orthographic consistency) and a significant effect of vocabulary matching (differences being larger in studies in which the groups were not matched on vocabulary knowledge). In regard to the former, a possible explanation might be that reading comprehension is a more demanding task in languages with low orthographic consistency (e.g., English, French) because decoding (one of the building blocks of reading comprehension according to the “simple view of reading”) is more challenging for children with dyslexia in languages with low orthographic consistency ….

To conclude, we found that individuals with dyslexia experience large difficulties in reading comprehension. The estimated effect size (g = 1.43) is not as large as that reported for word reading skills in previous meta-analyses on dyslexia (e.g., Melby-Lervåg et al., 2012; Parrila et al., 2020a) and this reinforces the notion of considering reading comprehension a secondary symptom of dyslexia. At the same time, we found a small effect on listening comprehension, which reinforces the finding of previous studies that children with dyslexia may experience deficits in broader language skills. Taken together, our findings suggest that the reading comprehension deficits in individuals with dyslexia are likely the product of deficits in both decoding and oral language skills.