Monday, January 23, 2012

Meta-analysis continues to confirm importance of phonemic awareness in reading...but...lets not overestimate the importance

Double click on image to enlarge the abstract.

Meta-analysis like this one are powerful studies that confirm the importance, in this case, for phonemic awareness (Ga-PC per CHC theory) for reading and dyslexia. However I believe that often the individual studies used as the bases of meta-analysis may overstate the importance of particular abilities due to the well-known problem of specification error in each studies design. I've blogged about this previously and won't take up bandwidth reiterating the importance of recognizing how specification error can cloud accurate interpretation of studies. The long and short of the issue is that manly reading and dyslexia studies during the past two decades have suffered from specification only including indicators of the "hot and sexy" ability constructs in reading research and failing to include measures of known abilities that are also important in reading.

As a result of the "missing variable" problem, individual studies and, in this case, a meta-analysis of studies, most likely overstates the importance of the selected abilities this case phonemic awareness. For example, a recent synthesis by McGrew & Wendling (2010) demonstrated that other abilities often not included in the extant reading disability research (e.g., processing speed; associative memory; lexical knowledge) are important. If measures of these important reading-related abilities had been included in the studies summarized by this meta-analysis most likely the magnitude (effect sizes) of phonemic awareness would be lowered.

Yes...phonemic awareness (Ga-PC) is clearly important...but I caution readers to take the magnitude of the importance with a grain of salt as it is most likely somewhat less if all important reading-related variables had been included in the studies that are combined in this review

- Posted using BlogPress from Kevin McGrew's iPad

No comments: