Validity of the PPVT as a Test of Language Ability with Lower SES Children

Abstract

The Peabody Picture Vocabulary Test (PPVT) is often used as a substitute for individual intelligence tests. Indeed, Dunn (1965) claims that the PPVT provides an estimate of S's verbal intelligence through measuring his hearing vocabulary, and the age and standard score norms provided in the test manual are expressed in terms of mental age scores and intelligence quotients. Because of this approach, most PPVT validity studies have been concerned with assessing the test's comparability wirh measures of intelligence. In general, these smdies have reported only low to moderate correlations, and several investigators (e.g., Shaw, Matthews & Kl@ve, 1966; Brown & Rice, 1967) have concluded that the PPVT has inadequate validity as a test of intelligence. This conclusion suggests that the PPVT may be more appropriately used to assess language ability rather than cognitive functioning generally. The present study was therefore designed to evaluate the validity of the PPVT as a test of language ability by examining i n relationship to rhe 1961 edition of the Illinois Test of Psycholinguistic A b ~ l ~ r ~ e s (ITPA). Ss were 81 lower SES children (Mca = 65.42 mo , SD = 6.06 mo.) attending first grade classes at eight NSW state schools. There were 41 girls and 40 boys in the sample, 40 of whom were white and 41 part-aboriginal. The PPVT (Form A) and the ITPA were individually administered to each S by the same examiner on separate days. The PPVT was always the first test to be administered. Product-moment correlation coefficients were computed using raw score data. The correlations between the PPVT and the 9 ITPA subtests were: auditory decoding, .49; visual decoding, .36; auditory-vocal association, .66; visual-motor association, .23; vocal encoding, .45; motor encoding, .35; auditory-vocal automatic, .58; auditory-vocal sequential, .32; visual-motor sequential, .54. The correlation between the PPVT and the ITPA total score was .74. Except for the visual-motor association subtest, all coefficients were significant ( p < .01). Three points arise from the above findings. (1) The correlation of the PPVT with the ITPA total score is considerably higher than any of the correlations with individual ITPA subtests. This suggests that the PPVT has value as a test of language ability generally rather than as a measure of a particular linguistic skill. ( 2 ) When correlations with ITPA subtests are examined, it is of interest to note that the PPVT is most closely associated with the auditory-vocal association and auditory-vocal automatic subtests. In common with the PPVT, both of these subtests require the comprehension of auditory stimuli. ( 3 ) The magnimde of the correlation with the ITPA total score suggests that the PPVT has reasonable validity as a test of language ability. The findings therefore support the contention that the PPVT may be more appropriately used to assess language ability, rather than intelligence, at least with lower SES children.