A five year analysis of factors affecting the retention of in-state freshmen admitted to the University of Tennessee from fall quarter of 1968 through summer quarter of 1973

The purpose of this study was to provide administrators, faculty, and staff at The University of Tennessee, Knoxville with information concerning the retention of in-state freshmen. Specific attention was focused on the analysis of the data concerning the retention of marginally qualified in-state freshmen. This study was designed to provide the historical information necessary to make future decisions on the value of and need for special assistance programming for students considered only marginally qualified for admissions.

It was necessary to use a variety of computer procedures to generate a master tape containing the information needed in this analysis. Basically, the computer information available on every instate freshmen admitted from Fall Quarter of 1968 through Summer Quarter of 1973 with a composite American College Test (ACT) score as part of their individual record was selected for analysis. Every individual identified on the file tape was "tracked" for a minimum of five full years or twenty consecutive academic quarters. It could thus be determined if a student entering in Summer Quarter of 1973 graduated within five full years or by the end of Spring Quarter 1978.

A cumulative five-year record was generated on every individual identified for analysis. The individual information available for analysis consisted of student identification number, quarter and year of enrollment, composite Act score, high school cumulative grade point average, home county, sex, marital status, race, date of birth, college of initial enrollment, university graduation date, cumulative college grade point average, and enrollment status for 20 consecutive quarters.

The resulting computer file tape consisted of individual information of 17,305 in-state freshmen from this five-year period. The sample selected through the process described was considered more than sufficient not only for making inferences about the total group or entire population of individuals originally admitted as in-state freshmen during this five-year period but also for conceptualizing the probable persistence patterns of in-state freshmen who have subsequently been admitted.

The data were analyzed descriptively by the computer providing frequency and percentage of individuals in each of the eight biographic or demographic categories: race, sex, marital status, composite ACT score, high school cumulative grade point average, county wealth, enrolled college and age. In addition, the descriptive analysis was further supplemented by comparatively graphing the persistence patterns for 20 consecutive quarters of the students in these same categories. Stepwise discriminant analysis was utilized as a classification technique for predicting the probability of graduation.

Six research questions were employed to design the analysis of the data and to present the findings.

1. The study found that 38.8 percent of all in-state freshmen admitted from Fall Quarter of 1968 through Summer Quarter of 1973 graduated within 20 consecutive quarters from their respective dates of admissions.

2. The research determined that of all individuals admitted during these five years as in-state freshmen, 97.3 percent were under 21 years of age, 91.2 percent were not married, 91.4 percent were white 84.5 percent had a high school grade point average of 2.25 or higher, 89.5 percent had a composite ACT score greater than 16, and 55.8 percent were males while the remaining 44.2 percent were females.

The research determined the comparative composition of those graduated within 20 consecutive quarters and found that 98.l percent were under 2l years of age at their respective dates of admissions, 89.6 percent were not married when admitted, 95.4 percent were white, 92.9 percent had a 2.25 or higher high school grade point average at date of admissions, 95.7 percent had better than a 16 composite ACT score, and 53.7 percent were males while the remaining 46.3 percent were females. When the percentages of those graduated within 20 consecutive quarters were compared with percentages admitted, negligable differences were found by county wealth and college enrollment variables.

Of all individuals admitted with a composite ACT score of 16 or less, only 19.5 percent persisted to graduation. Also, of all individuals admitted with both a composite ACT score of 16 or less and a high school grade point average of less than a 2.25, only 9.2 percent persisted to graduation within 20 consecutive quarters.

3. The research determined that the greatest attrition of the students admitted as in-state freshmen occurred during the first academic year. The largest portion of the students were graduated on schedule, four academic years or 15 academic quarters from their respective dates of admissions.

4. Concerning persistence and graduation patterns, females tended to be graduated earlier or exactly on schedule proportionately more frequently than did males.

vi Individuals originally enrolled in the College of Liberal Arts and the College of Education tended to graduate on schedule more frequently than students in other colleges. Students originally enrolled in the College of Business took relatively longer to graduate than students originally enrolled in any other college. Virtually no students graduated ahead of schedule in the College of Engineering; a few were graduated during the fourth academic year, but most individuals graduated exactly on schedule at the end of the fifth academic year in accordance with that college's five-year programs.

The rather homogenous composition of the in-state freshmen complicated the comparison of persistence and graduation patterns on the basis of age, marital status, county wealth, high school grade point average, composite Act score, and race.

5. The research determined that composite ACT score and high school grade point average were both positively and significantly related to graduation while the county wealth variable did not significantly improve graduation predictions. High school grade point average proved to be the single best predictor of graduation from The University of Tennessee, Knoxville; however, composite ACT score did add significantly to the graduation prediction. Utilizing only these two variables, it was possible to correctly classify on an individual basis 63.4 percent of all in-state freshmen as future graduates or non-graduates at their respective dates of admissions.

6. The research determined that 56.0 percent of those students not graduated within five years had better than a 2.00 cumulative grade point average at the time they left The University of Tennessee, Knoxville.