On the sixteen campuses of the University of North Carolina, student retention is becoming a strategically critical factor. For the period from 1998 to 2007, the University of North Carolina System had an average campus freshman retention rate of 80.8%. The North Carolina University System is composed of sixteen individual, public institutions located throughout the state. The current economic situation prompted an examination of student retention and graduation rates at each of the System's sixteen institutions. Administrators at each institution have been charged to review the status of student retention at their institution and to take the necessary measures to improve retention rates and to set retention goals. Taxpayers want to know their tax dollars are well spent. In North Carolina, retention of students at the University of North Carolina campuses is of utmost importance for institutional administrators to ensure efficient stewardship of taxpayers' funds. Retention of students is becoming a fundamental consideration for all university faculty and staff at North Carolina higher education institutions. Due to the generally accepted higher costs of their programs, engineering and technology faculty and administrators should be especially cognizant of retention rates, and the relationship of SAT scores to the successful completion of their programs by their students. A linear regression model to predict an expected campus freshman retention rate was produced using average campus SAT scores. Initial examination indicates that some campuses have lower average freshman retention rates than other campuses due substantially to the level of freshman high school preparation as reflected by Scholastic Aptitude Test (SAT) scores. Student performance on the SAT is a primary indicator of first year student success, but other factors may also influence retention and provide opportunities for faculty and administrators to improve retention rates above expected levels as predicted by the SAT alone. After comparing SAT scores and retention rates at the sixteen state institutions, a regression model was built to predict expected retention rates which were compared to recent historical retention rates. Recommendations to improve retention rates are provided.
