Visuoperceptual-orthographic reading abilities: A confirmatory factor analysis study

Abstract

The current study examined the psychometric properties of a hypothetical visuoperceptual-orthographic (VPO) reading construct in a sample of nonimpaired college students (N = 152). Participants were administered a battery of standardized and experimental measures of VPO, including Woodcock–Johnson III (WJ-III) Letter–Word Identification (Woodcock, McGrew, & Mather, 2001), Visual-Naming Speed (word page of the Stroop), Letter-Identification Task, Same/Different Letter Decision Task, Word Matching Task, Homophone Decision Task, Pseudohomophone Decision Task, and Word Jumble Task. The LISREL 8.54 computer program (Jöreskog & Sörbom, 2003) was employed for confirmatory factor analysis to assess the tenability of a multifaceted, unitary VPO reading construct. Goodness-of-fit statistics indicated that VPO is a hierarchically organized construct with one 2nd-order factor (i.e., VPO) and three 1st-order factors (i.e., perceptual processing speed, prelexical accuracy, and lexicosemantic accuracy). Alternative models were tested but produced unsatisfactory goodness-of-fit statistics. Altogether these findings are in agreement with those of previous cognitive and neuroscientific studies and further support the notion that VPO should be viewed as a unique factor in the assessment, diagnosis, and remediation of developmental dyslexia.

This research was conducted as partial fulfillment of the requirements for the first author's Master of Science degree at the University of Wisconsin–Milwaukee. A portion of these data was presented at the International Neuropsychological Society, February, 2007, Portland, OR.

Notes

¹A recent exploratory factor analysis was conducted in our laboratory, including all of the current experimental tasks (Mano, Osmon, & Klein, 2005). Results from these analyses were reported at the National Academy of Neuropsychology, Tampa, FL, USA. Variables associated with morphological awareness were excluded from the current analysis because they possessed poor communality scores.

²Such early perceptual processes are perhaps best indexed by inspection time, a paradigm that has been shown to correlate with broad reading ability (Grabe, 1981; Whyte, Curry, & Hale, 1985).