Computerised adaptive testing

Since the early 1990s, educators have begun to recognize the potential of computer technology to advance the application and development of adaptive testing procedures. Reckase (1988) defines computer adaptive testing (CAT) as a computerized testing procedure that selects items to match the ability of the examinee during test administration. Most CAT procedures are designed to operate relatively autonomously of tester intervention. Thus, the computer selects and presents test items appropriate for the examinee’s ability level, the examinee responds to these via the computer keyboard or mouse, and results are processed automatically on completion of each test item. The CAT process typically involves two main steps (Weiss, 1983). In the first, an initial item is selected whose difficulty level matches the examinee’s current estimated ability (Wainer, 2000). In the second, the examinee’s responses to this item are scored, and their ability estimate is updated to incorporate this information. If the examinee answers the item correctly, a more difficult item will be selected at the next step. An incorrect response will prompt the selection of an easier item from the pool.