ABSTRACT
The computer is poised to become the preferred mode by which large-scale tests are administered. The advantages of testing by computer are well-documented and have been realized across a diverse range of assessment settings, including certification and licensure, employment, clinical, higher education admissions, and K–12 education (Parshall, Spray, Kalohn, & Davey, 2002; Puhan, Boughton, & Kim, 2007). At first consideration, testing by computer poses no obvious concerns for standard setting, as most standard setting methods apply equally well to computer-based or traditional paper-based testing programs. However, several aspects of computer-based assessment raise issues for standard setting practice. For one, the opportunities to assess in innovative ways using computer technology could pose challenges for many of the standard setting approaches used today. But an even more salient issue that can significantly complicate standard setting with computer-based assessments is dual-mode testing; that is, when a testing program administers both computer-based and paper-based versions of the same assessment at the same time. In this case, many of the concerns about comparability of scores across testing mode may also affect the standard setting process.
