Consrruct Validity of the Illinois Test of Psycholinguistic Abilities

Publisher Website
Google Scholar
Abstract
The construct validity of the ITPA was investigated by factor analyzing the subtests with 20 criterion tests that were designed either to match the subtests on each psycholinguistic variable but vary in content, or to differ in channel of communication but match in all other respects. Construct validity was determined by the degree to which each subtest measured discrete, independent abilities and the extent to which these abilities represented the constructs of the Osgood model of language behavior. Results generally confirmed the construct validity of the test. Nine of the subtests measured discrete abilities, and two dimensions of the underlying model, i.e., level and process, were substantiated. The channel of communication, particularly the visual modality, constitutes the greatest threat to validity.