Convergent and Discriminant Validation of the Trait of Visual Figure-Ground Perception

Abstract

The study investigated the construct validity of the trait of visual figure-ground perception by means of a multitrait-multimethod analysis of visual figure-ground perception, form recognition, and verbal ability, as well as the use of four methods of measuring those specific traits. Subjects were 64 upper-middle-class first grade children, mean CA 6-7 years, who were screened for intelligence (Slosson IQ =99.9) and to rule out visual perceptual problems. Measures of the trait of visual figure-ground perception were found to possess a moderate degree of convergent validity but discriminant validity was not demonstrated. Considerable method-of-measurement bias and shared trait variance between visual figure-ground perception and form recognition were found to exist. Construct validity was not demonstrated for the trait of visual figure-ground perception as a distinct visual perceptual skill. Cautions are presented for the educational use of “visual figure-ground perception.”