ABSTRACT
Among several recent trends in educational measurement is one toward providing information that more directly benefits individuals as opposed to institutions. One instance of such measurement is diagnostic assessment (e.g., Bejar, 1984; Forehand, 1987), which attempts to offer guidance to teachers and students about the specific problems encountered in learning and how they might be addressed. A second, related instance is constructed-response testing (e.g., Ward, Frederiksen, & Carlson, 1980), in which the test task is closer to that required of students and workers in academic and vocational settings, and therefore of more obvious relevance to the examinee. Both forms of assessment have been practiced for many years; diagnostic assessment has long been a standard, though typically informal, practice of master teachers and constructed response assessment a routine component of classroom tests, graduate school comprehensive exams, and even some standardized testing programs (e.g., the Advanced Placement Program of the College Board).
