On Examinee Choice in Educational Testing

Abstract

Throughout our society, individuals are compared on summaries of diverse measures. Often the measures are neither the same for all competitors nor selected randomly from a universe of measures. In fact, comparisons often are based on a mixture of measures in which the competitors choose some or all of the elements of the mixture; applicants to colleges choose some of the courses they take in high school, which extracurricular activities they participate in, and (to some extent) which entrance exams they take. What are the consequences of any selection procedure, if such choices are allowed? How can the adverse consequences be minimized? This article summarizes results from tests that have allowed examinee choice heretofore and provides a basis for a fuller discussion of these issues.