Self-Adapted Testing

Abstract

In self-adapted testing (SAT), examinees choose the difficulty of each item they attempt immediately before it is presented. In this article, I review the rationale for investigating SAT and research exploring the effects of SAT on ability estimates, precision and efficiency, mechanisms underlying the effects associated with SAT, and examinee reactions to SAT. Some of the research demonstrates that SAT leads to higher ability estimates than computerizedadaptive testing (CAT), whereas other research demonstrates no main effect of test administration mode but that SAT leads to ability estimates that are less influenced by extraneous attributes of the examinee (e.g., test anxiety). SAT is less efficient than CAT, but more efficient than fixed-item testing. Examinees strongly endorse the main elements of SAT.