Kim, D. H., & Huynh, H. (2010). Equivalence of paper-and-pencil and online administration modes of the statewide English test for students with and without disabilities . Educational Assessment , 15 (2), 107–121. https://doi.org/10.1080/10627197.2010.491066

Achievement scores on paper-and-pencil and computer-based versions of a statewide test in English were compared.

Extant data from among the unspecified U.S. state's assessment scores of 31,651 students in grade 9 were studied; 7,000 matched pairs of students without disabilities and 483 matched pairs of students with learning disabilities were sampled for analysis.

The students' scores on the end-of-course English test in grade 9 was the dependent variable studied. Item level analysis and content bundle analysis were both performed. The five content areas included: Reading I–Understanding and Using Literary Texts, Reading II–Understanding and Using Informational Texts, Reading III–Building Vocabulary, Writing–Developing Written Communications, and Researching–Applying the Skills of Inquiry and Oral Communication.

First of all, the students without disabilities on average scored slightly better on the paper-and-pencil form, yet still at a level of statistical significance. The students with learning disabilities scored similarly on both test forms, with no statistically significant difference in mean scores. Some differences in the content areas, with small effect sizes, were noted. There were no differences in difficulty level for the online and paper-and-pencil groups for either the students with or without learning disabilities. Finally, there were none to mild DIF results on the item-level analyses; however, at the bundle (content area) level, the DIF results were significant but minimal in magnitude. In summary, the test administration modes demonstrated comparable results for both groups—that is, students without disabilities scored similarly on both online and paper versions of the test, as did students with learning disabilities.