ABSTRACT
The emergence and widespread adoption of item response theory (IRT) for test development, scoring, and equating have fundamentally changed testing practices since the nearly overwhelming number of publications of papers in the 1970s and 1980s. It is difficult, too, to imagine today's landscape of adaptive testing without IRT item parameters, models, and proficiency estimates, the latter computed “on-the-fly” as examinees sequence through a test. At the same time, however, IRT can also provide a valuable mechanism for communicating score meaning, by linking items to test performance to more directly illustrate how a given point on the score scale may be understood or interpreted in the context of (i) the contain domain of interest that a test item measures and (ii) what examinees know and can do. While much of the research connecting IRT with score reporting is generally concentrated in a few specific testing programs and interpretive contexts, the approaches described in this chapter have had considerable application across testing programs.
