Grady, L. B. (1992). Convergent and discriminant validation of four instruments assessing perceptual learning styles using a multi trait-multi method matrix approach (convergent validation) (Doctoral dissertation, University of South Florida, Tampa, 1992). Dissertation Abstracts International, 53(12A), 4186.
The purpose of this study was to validate the constructs of perceptual learning style through examining evidence of convergent and discriminant validity of four assessments designed to measure auditory, visual, print, and kinesthetic/haptic perceptual learning styles. The four instruments included were the CITE Learning Styles Inventory, Barbe & Milone's Find Your Learning Styles, the Edmonds Learning Style Identification Exercise (ELSIE), and the Multi-Modal Paired-Associates Learning Test Revised (MMPALT II). The four instruments used three different methods to assess perceptual learning styles: self-report, performance, and imagery. A multitrait-multimethod matrix was used to establish evidence of convergent and discriminant validation of the subtests measuring perceptual learning style traits. To accomplish this, the four instruments were administered to a group of 100 vocational-technical adult students who volunteered to participate in the project. This was not a random sample. Final results of the study are, therefore, only representative of this study group population. Results were correlated and organized in a multitrait-multimethod matrix. The matrix, using criteria developed by Campbell and Fiske (1959), revealed only one of 27 correlation coefficients was significantly different from zero p < .01 which would indicate convergent validity. This significant finding could have occured purely by chance. Continued evaluation of discriminant validity of this significant correlation coefficient revealed that of four comparisons made, only two were significant. The findings indicate a lack of discriminant validity. This study provided no supporting evidence for the validity of the constructs of perceptual learning style or for the four measures used to assess perceptual learning style. The major implication is that use of these instruments to measure perceptual learning style traits should be guarded, due to the possible invalidity of either the measures or the constructs of perceptual learning style.