Published by Research and Development Division American College Testing Program P. O. Box 168, lowa City, Iowa 52240 Summary This investigation compared the predictive validity of the ACT tests with the SAT and SCAT tests for a sample of 21 colleges and universities. Grades in specific courses as well as over-all grade point average were studied. Multiple R's were computed for each test battery. The ACT and SAT tests possessed about the same degree of predictive validity, and both were better predictors than the SCAT. Predictive results varied from school to school and from course subject to course subject, thereby suggesting that predictive validity for individual colleges and universities and for specific subject matter areas should be established. In addition, a college's choice between the ACT or SAT probably can not be made on the basis of their relative predictive validities, but rather should be made on the basis of the total programs of services offered by the respective testing agencies. Comparative Predictive Validities of the American College Tests and Two Other Scholastic Aptitude Tests Leo Munday Standardized tests of academic potential are used in nearly all American colleges and universities for admissions and guidance purposes. Institutions have a wide variety of such tests from which to choose. While many considerations are relevant to this choice, a major consideration is predictive validity. This report compares the predictive validities of three of the most widely used tests of scholastic potential. These were the American College Testing Program's tests (ACT), the College Entrance Examination Board's Scholastic Aptitude Test (SAT), and the Educational Testing Service's School and College Ability Test (SCAT). The ACT battery consists of tests in English, mathematics, social studies, and natural science; it requires 180 minutes. Both the SAT and the SCAT have verbal and quantitative parts; the SAT requires 180 minutes and the SCAT requires 70 minutes. Procedure Sample. Both ACT and SAT scores were available for freshmen at five colleges and universities which participated in the 1964 ACT Research Service. ACT and SCAT scores were also available for students at six other colleges and universities. Additional data from three other studies were made available to the investigator. In two of these (Chase, et al., 1963, Klingelhofer, 1964), the ACT and SAT tests were administered to randomly selected and universities. Since some of the colleges subdivided their samples by sex or curriculum, we obtained a total of 29 groups from the 21 colleges. All tests were taken prior to college instruction, and all grades were earned during the freshman year. Table 1 describes these institutions in terms of their enrollment, control, level of instruction, and geographic region. |