The Equivalence of Paper-and-Pencil and Computer-Based Testing

This is the authors' abstract. We don't add key points for this paper.

In recent years, there has been increased interest in converting from paper-and-pencil to computer-based tests. As more and more organizations consider converting from paper-and-pencil tests to computer-administered ones, the importance of such research increases dramatically. This article examines the research showing the equivalence of paper-and-pencil–administered tests and computer-administered tests, especially when the computer-administered test is exactly the same as the paper-and-pencil test. Four different categories of research on testing by computer are examined: previous reviews of research, standards and guidelines for computerized testing, individual research studies, and studies about computerized testing over time. From this review, a number of general conclusions about testing by computer are reached: (a) it can be equivalent to a paper-and-pencil test, but it is the responsibility of the test developer to prove that it is; (b) this equivalence is established by meeting stringent criteria; (c) computers affect testing; (d) special considerations are necessary when computers are used for tests; and (e) test users must have at least a basic understanding of the computer and its vicissitudes, in conjunction with a knowledge of the psychometric properties of the test, to effectively use computer testing.