The construct validity of two versions of a Preschool Vigilance Test
Abstract
The purpose of this study was to assess the construct validity of two versions of a Preschool Vigilance Test (PVT). One version of the PVT consisted of interstimulus intervals ranging from 10 to 60 seconds and the other version consisted of interstimulus intervals ranging from 5 to 30 seconds. The construct validity of five measures obtained from each version of the PVT was assessed. Based on theories of vigilance and previous research it was proposed that both total Errors of Omission and Omission Ratio scores reflect Inattentiveness, Errors of Commission reflect Impulsiveness, Nonorients reflects Hyperactivity, and Reaction Time reflects Efficiency. In order to assess the construct validity of these measures the multitrait-multimethod approach was employed. This approach involved correlating children's performance on the vigilance measures with other measures of inattentiveness, impulsivity, hyperactivity, and efficiency. It was expected that correlations between two different methods assessing the same trait would be significantly different from zero and thus support the convergent validity of the methods. It was also expected that correlations for the same method, different traits would not be significantly different from zero and thus support the discriminant validity of the method. Both convergent and discriminant validity were considered necessary to support the construct validity of the PVT. Results indicated that only two of the ten and in no case were the convergent validity values greater than all other correlations in the multitrait-multimethod matrices. These results, therefore, did not support the construct validity of the PVT as had been proposed. Based on the results of this study, it is proposed that rather than the different PVT measures reflecting separate constructs, a combination of the PVT measures may reflect Inattentiveness. This study also examined the effects of children's age and gender on vigilance performance. Results indicated that older children received scores indicating greater accuracy in stimulus recognition than younger children. However, no significant differences in vigilance performance were found between boys and girls. Finally, the relationship between vigilance performance and IQ was examined. Results indicated that IQ scores were significantly correlated with several PVT measures. The need for further research on the PVT is discussed.
Degree
Ph.D.
Advisors
Ottinger, Purdue University.
Subject Area
Psychotherapy|Developmental psychology|Psychological tests
Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server.