The predictive validity of the Short test of educational ability for three Virginia community colleges

TR Number



Journal Title

Journal ISSN

Volume Title


Virginia Polytechnic Institute and State University


Community college counselors are continually confronted with the difficult task of assisting students to choose from one of the various curricula. Scholastic aptitude test data have been frequently used by counselors to provide students with information to assist them in their college planning. However, before the decision to assess the scholastic aptitude of students and to use the results for counseling purposes, the scholastic test's predictive validity should be determined.

The Snort Test of Educational Ability (STEA) is a scholastic aptitude test score available to all Virginia Community Colleges. The STEA score is currently transmitted to the college via the student's high school transcript. Because the 1975 Virginia Community College System Policy Manual requires a validity study of potential curricular placement tools, it was the purpose of this study to determine the STE.A's validity for predicting a student's 1976 fall quarter grade point average.

This investigation was conducted during the academic year 1976-1977 at three community colleges within the Virginia Community College System. The samples used in this study were drawn from 1976 Virginia high school graduates enrolled for the first time for the fall quarter, 1976, at a metropolitan, a suburban, and a rural institution. One hundred subjects were randomly selected at each institution. From the colleges' permanent files the following data were collected for each student: (1) total high school grade point average, (2) last high school grade in English, (3) last high school grade in math, (4) STEA score, (5) sex, (6) race, (7) curriculum, and (8) 1976 fall quarter grade point average. Using this data, Pearson product moment correlations and multiple regression coefficients were generated by the"Correlational Procedure" and"General Linear Model Procedure” of the Statistical Analysis System - 1976

The results of the study as they related to the hypotheses were as follows. At the three institutions significant (p< .05), but low, correlations were found between the students’ fall quarter g.p.a. 's and their STEA scores. Low to moderate three and four variable multiple correlations were significant (p < .05) between the fall quarter g.p.a. and combinations of the following independent variables: the total high school grade point average, the grades in high school English and math, and the STEA. For all three colleges the results were not significant for the prediction of the fall quarter g.p.a. using the STEA when subgrouped by students’ sex, race or curriculum.

From the results of this study the following conclusions were made. For the three com.mur.i