| As a growing number of states have administered the same tests in both paper-and-pencil testing (PPT) and computer-based testing (CBT) formats simultaneously, the question has been raised whether or not scores obtained from the two administration modes are equivalent or interchangeable. The purpose of this study was to compare student performance between PPT and CBT of large-scale statewide end-of-course examinations in the four subject areas of Algebra, English, Biology, and Physical Science. Overall, the results at both the item level and the test level support the comparability of PPT and CBT. Although some differences were observed, the overall results suggest that these differences were relatively small. For all subject areas, the model fit the data adequately for both PPT and CBT. Although there was some lack of invariance, fit indices showed that all three models of equivalence fit the data adequately and the additional restrictions did not adversely affect model fit for the four subject areas of Algebra, English, Biology, and Physical Science. No evidence was found to support the claim that the effects of the number of items tied to a common reading passage in the English test on students' performances varied across the administration modes. |
