CONVERGENT AND ECOLOGICAL VALIDITY OF THE WOODCOCK JOHNSON PASSAGE COMPREHENSION TEST AND THE WECHSLER INDIVIDUAL ACHIEVEMENT TEST- READING COMPREHENSION SUBTEST

Author

Maria Stacy, Southern Illinois University CarbondaleFollow

Date of Award

12-1-2021

Degree Name

Doctor of Philosophy

Department

Psychology

First Advisor

Kibby, Michelle

Abstract

Reading comprehension assessments often vary from one measure to another related to the response format required, passage length and other variables. Yet, these measures purport to assess the same skill of reading comprehension, and they are often used interchangeably. Over the last decade some reading researchers have raised concerns that the variability in reading comprehension assessments may mean these measures are assessing different components of reading comprehension instead of consistently assessing one well-defined, complex skill. This study compared two commonly used reading comprehension assessments: the Wechsler Individual Achievement Test- Reading Comprehension (WIAT-RC) subtest, which requires examinees to read longer passages and answer open ended questions, and the Woodcock Johnson Passage Comprehension test (WJ-PC), which requires examinees to read very short passages and fill in the one word that is missing. This study was designed to test whether performance on these two measures is predicted by the same underlying language and executive functioning skills (a form of convergent validity) and whether these two measures are commensurate in predicting performance on a ‘real-world’ reading comprehension task (GRE verbal questions), with the latter being a test of ecological validity. Results suggested that the two measures varied in their ability to predict performance on the GRE verbal questions, where the WJ-PC significantly predicted performance on these questions, while the WIAT-RC did not. None of the executive functioning skills included in analyses significantly predicted performance on either the WJ-PC or the WIAT-RC, so no difference was found there; however, there were differences in how the language skills predicted performance on the two measures. Both vocabulary and word reading predicted significantly more variance in performance on the WJ-PC than the WIAT-RC. Overall, these results suggest there are important differences between these two reading comprehension measures, related to which underlying skills influence performance on each measure and the two tests’ ecological validity. These differences raise concerns about how well the reading comprehension construct has been defined and how consistently that complex skill is assessed across different measures. In addition, the differences between these two tests should be considered by clinicians who use and interpret scores on these measures in clinical settings.