problems associated with the analysis of gain scores, residual scores, and repeated measures for individuals is clear and logical. However, his choice of ACT's COMP Objective Test to illustrate his argument is misleading. ACT formally notified Pike and his colleagues of the inappropriate analysis of data based on this test in Fall 1989 (Steele, 1989c). At that time some of the same problems he discusses in his article were pointed out as leading him to incorrect results and unwarranted generalizations. Unworthy uses of tests should not be used to judge the technical merits of those tests. The Standards for Educational and Psychological Testing (AERA, APA, and NCME, 1985) make clear that validity is not a characteristic of an instrument per se but "refers to the appropriateness, meaningf ulness, and usefulness of the specific inferences made from test scores" (p. 9). Pike claims inaccurately that, "Gain scores are reported by the College Outcome Measures Program (COMP) staff as indicators of student learning and program effectiveness" (p. 72). In fact, ACT has always recommended mean score gains, not individual gain scores, as indicators of program effectiveness. In addition, ACT has always clearly stated that the COMP Objective Test should not be used in making judgments about individual students but should be analyzed with the group as the unit of analysis as an indicator of program quality and effectiveness (Steele, 1989b). The merit of the COMP Objective Test can be judged against accepted technical standards, based on its intended uses. The worth of the Objective Test depends on actual uses that do not undermine the construct validity of the instrument and represent well-designed evaluation efforts (Scriven, 1981). Pike describes results from his unworthy use of the test to question its technical merit. In contrast to the somewhat meaningless results based on his use and