The construction of scales for predicting academic success for freshman students at West Georgia College
Students entering college need effective predictive information in order to evaluate their likelihoods of success thereby allowing them to establish determination to function at a sufficient level of motivation to be successful in college. Such predictions of the chances of success also provide the institution with useful information when decisions become necessary in determining allocation of quarters, aid, and even in determining admissions when applications exceed facilities. It was the purpose of this study to provide scales which would yield this same such information to the students and administration of the West Georgia College located in Carrollton, Georgia. To that end a predictive technique, first used by Glueck and Glueck (1940) to predict recidivism of delinquent behavior among juveniles and later by Ford (1949) to predict various aspects of academic attainment (1949) to predict various aspects of academic attainment at the high school level, was employed in the present study. Success was defined in this study as a freshman average grade of 2.1 or better. Letter grades were awarded the following numerical equivalents for the purpose of computing freshman average grades: A - 4 points; B - 3 points; C - 2 points, D - 1 point; and F - 0 points. The sample group consisted of 341 students for whom Scholastic Aptitude Test - Verbal, Scholastic Aptitude Test - Math; Scholastic Aptitude Test - Total, and Freshman Average Grades were available. The records described were found among the records of 402 entering freshmen in the Fall Quarter of 1960 at West Georgia College. The three Scholastic Aptitude Test Scores were used to establish predictive tables by test score and to provide point values to be used in computing total predictive scores. Freshman average grades were used to distinguish between successful and unsuccessful within each class interval in establishing frequency distributions. The method used was as follows: 1. Frequency distributions were made for each of the three test scores, locating each student in an interval according to his score and in the appropriate successful- unsuccessful column for that interval. 2. The percent of successful students was calculated for each interval. 3. These percent scores comprised the score sheet (Table 1). 4. These percent scores were treatedas raw points, were totaled, and became total predictive score. 5. The resulting total predictive scores were then tallied, a frequency distribution was made with both successful and unsuccessful frequencies, and Table 2 resulted. Percent of successful students was calculated for each interval thus providing chances in 100 of success for each interval based on the total predictive score. The scale was then validated by using 341 students' scores gleaned from entering freshmen in the Fall quarter of 1961 using the same criteria used for the sample group. The method for validation was as follows: 1. Students were divided according to freshman average grades into successful and unsuccessful groups with an average of 2.1 or above considered successful. 2. Total predictive scores were computed for each using date from Table 2. 3. A frequency distribution was made which had successful- unsuccessful column as did Table 2. The rationale for validation was to make a gross prediction of success. That is to say that if there existed greater than 50% chance of success for cases in any interval according to percentages found in Table 2, then it was predicted that all students whose scores fell into that interval would be successful. Conversely, if scores fell into intervals that predicted less than 50% successful then all were predicted to be unsuccessful. The results were that a gross prediction of success was made for all students whose total predictive scores were 151 or greater and students whose total predictive scores were 150 or less were predicted to be unsuccessful. Compared on the basis of these gross predictions, the accuracy of prediction within the definition of success was 62.7 percent, accuracy of prediction for unsuccessful was 64.1 percent, and the accuracy of prediction for the total validation group was 63.3 percent. Accuracy of prediction for separate intervals ranged from 0 percent to 100 percent. It should be noted, however, that the intervals which yielded 0 percent and 100 percent contained one and zero students respectively.