The new assessment of orthographic processing

The new assessment of orthographic processing

First Author: YI-JUI Iva CHEN -- Rightpath Research & Innovation Center. University of South Florida
Keywords: Assessment, Diagnostic measure, Orthographic Knowledge, preliteracy skills
Abstract / Summary: 

Two studies were conducted to test the reliability and validity of a computer-based orthographic processing assessment that pinpoints an individual’s ability to perceive, store, retrieve, add, remove, and arrange orthographic representations (consisting of English letters). This assessment is comprised of six different tasks. Each task has 9 to 12 actual items plus 1 to 4 practice items. The first five tasks require a fixed response, and the sixth task requires an open-ended response.
In the first study, the analyses were based on 30 American elementary school students demonstrated high reliability: Cronbach’s alpha was 0.92, Pearson’s separation was 0.89, Spearman-Brown’s correlation to the alternative form was 0.94. Spelling and word recognition ability were used as external variables to test validity. The data showed that the assessment had a strong correlation (r = 0.86) with the elementary spelling inventory (Bear et al., 2000) and a strong correlation (r= 0.84) with the San Diego Quick Reading Assessment (McKenna & Stahl, 2008).
In the second study, the analyses based on 34 English as foreign language elementary school students in Taiwan showed high reliability: Cronbach’s alpha was 0.87, Pearson’s separation was 0.84, and Spearman-Brown’s correlation to the alternative form was 0.82.
The Letter Identification and Word Identification subtests of the WJ-III and a prior orthographic knowledge task (Olson et al.,1985) were used as external variables, both with a moderate to strong correlation (r=0.79, r=0.73 respectively). Moreover, the Elision subtest of the CTOPP was used as an additional external variable. As found by previous studies, the orthographic processing assessment positively correlated (r=0.65) with phonological processing, but the degree of the correlation was weaker than the correlation with the tasks involving orthographic representations.