THE PREDICTIVE VALIDITY OF THE AMERICAN COLLEGE TEST FOR STUDENTS FROM

scores can be used to foretell college grades of students from the general population, few efforts have been made to assess the pre-dictive validity of the ACT for persons from specific socioeconomic backgrounds. Since students from low income families on the average do make lower scores on the ACT than do persons from other socioeconomic backgrounds (Merritt, 1970), the ACT may be less valid as a predictor of their college academic success. The purpose of this study, therefore, was to determine predictive va-lidity of the ACT for students from low socioeconomic levels. Method. The sample for this study consisted of individuals employed on the Federal Work-Study Program during the school sessions of 1968-69 and 1970-71. All members of the sample earned a minimum of $50 during the fall semester by working on this labor program, had a composite ACT score on record with the college, and lived in either dormitory facilities or apartments for married students. The Work-Study Program eligibility re-quirements assured that all were from low income backgrounds. A regression line between fall semester GPA and composite ACT score was computed for the 204 persons enrolled during 1968-69