To take advantage of computers as a platform, it is natural for test developers to convert their existing paper-and-pencil tests and administer them via this medium. Much of the early work on computer-based testing was focused on converting paper-and-pencil tests such as the California Psychological Inventory (CPI; Scissons, 1976), Sixteen Personality Factor (16PF; Harrell & Lombardo, 1985), and Job Descriptive Inventory (JDI; Donovan, Drasgow, & Probst, 2000) to computer versions.
Converting tests to a computer-based administration has several advantages. Once a test has been developed for computer administration, the cost of administration can be relatively small. There is no longer a need to print and transport paper forms, the computer can administer the test and consequently a proctor may be unnecessary for some types of assessments, and tests can be scored quickly and accurately. Moreover, some problems such as missing responses can be prevented by the test software, and administration errors by fallible proctors can be minimized. Examinees' response times for each item can be tracked and used to identify some types of testing problems. Additionally, with computer-based scoring, it is possible to provide instant feedback regarding an individual's performance on a test.
However, converting a test from a paper-and-pencil format to a computer format is not without difficulties. Green, Bock, Humphreys, Linn, and Reckase (1984) noted that computerized tests cannot be automatically assumed to yield scores that are comparable to their paper-and-pencil counterparts: "The two tests are equally valid only if they have been demonstrated to yield equivalent meas-
Was this article helpful?