Computer Adaptive Testing (CAT)
W. Holmes Finch, Brian F. French · Educational and Psychological Measurement · 2018
This chapter explains the linear testing as the administration strategy where an individual completes a full set of test items prepared prior to the administration, and not geared for an examinee&s;s particular ability level on the construct of interest. It describes the core elements of Computer adaptive testing (CAT) design, and the advantages and disadvantages of its use. The chapter outlines issues around the selection of items, control of item exposure, standards for ensuring that construct content is well-covered by the instrument, and selection of items to optimize estimation of the ability parameter. It also discusses CAT algorithms for which individual items are selected at each stage of the test administration. Selection of the item that maximizes information, given the value of the latent trait estimate, is probably the most commonly used approach in CAT. CAT offers a number of advantages when compared with the standard paper and pencil test format.