An Evaluation of the Validity of the Test of Visual Perceptual Skills - Revised (TVPS-R) Using the Rasch Measurement Model

Abstract

The visual perceptual skills of children are often evaluated by health care and education practitioners. Even though the Test of Visual Perceptual Skills - Revised (TVPS-R) is one of the most frequently used instruments with school-age children, its construct validity has not been evaluated thoroughly. The purpose of the study was to evaluate the scalability/interval level measurement, unidimensionality, lack of differential item functioning (DIF), and hierarchical ordering of items of the TVPS-R and its seven subscales using the Rasch Measurement Model (RMM). The TVPS-R scores from a sample of 356 normally developing children (171 boys and 185 girls), ranging in age from 5 to 11 years, were used to complete the RMM analysis. When the seven individual TVPS-R scales were analysed, they all exhibited adequate measurement properties (scalability/interval level measurement, unidimensionality, lack of DIF, and hierarchical ordering). However, when they were collapsed together to form an overall composite scale of motor-free visual perceptual skills, the TVPS-R items failed to group together to measure a unidimensional construct. In addition, many scale items exhibited RMM misfit or DIF. The results suggest that the seven TVPS-R subscales can be used on an individual basis with clients to generate a profile of their motor-free visual perceptual skills, but that they cannot be summed together to calculate an overall summary motor-free visual perceptual score or perceptual quotient. The TVPS-R composite scale does not exhibit adequate construct validity.