Parental contributions to longitudinal stability of cognitive measures in the Colorado Family Reading Study.

Data on reading-disabled and control probands and their families tested in the Colorado Family Reading Study were subjected to a comprehensive longitudinal-familial path analysis. Covariance matrices were constructed from composite scores for reading performance and for coding and symbol processing speed, and maximum likelihood estimates of familial parameters were obtained. With regard to the reading data, these estimates suggest that parental influences contribute importantly to longitudinal stability for reading-disabled children but not for controls. Such a difference in covariance structure between the 2 types of families could be due to a gene (or genes) with major effect. In the case of coding and symbol processing speed, covariance structure is similar for the 2 types of families, and combined parameter estimates indicate that parental contributions to longitudinal stability for this measure are relatively unimportant.