Assumptions About True-Scores and Populations in Equating

摘要

This excellent set of papers is comprehensive and very well written. The Kane et al. paper lays out the theory for linear equating with the NEAT design using a clever but simple framework. The Suh et al. paper is an excellent empirical study of the various methods. The Mroch et al. paper provides an insightful evaluation of the methods as developed and studied in the other two papers. Although the scope of these papers is broad in the sense of the methods covered, the scope is somewhat restricted in two senses. First, results are provided for an internal anchor only. Second, if a synthetic population is involved, the weights are always w_1 - 1 and w_2 = 0. The latter issue is discussed more fully after considering some comments about true-score assumptions. My comments are intended to be thought provoking, not negative.