highly significantp < 0.001
As shown in Table 7 , all models were highly significant ( p < 0.001) for the prediction of the following language measures: Boston Naming Test , oral reading of single words, written spelling to dictation, as well as oral and written naming of the objects from the WAB.