Noncognitive predictors of academic success for first generation college students at a public research university in Texas

In 1995-1996, first generation college students accounted for 53% of the population in two-year colleges and 34% for four-year institutions (Choy, 2001). Although these numbers are promising in terms of access, baccalaureate attainment rates are far less encouraging. According to data from the U.S. Department of Education’s 1988 National Education Longitudinal Study Postsecondary Education Transcript Study, only 24% of first generation college students who graduated from high school in 1992 who enrolled in higher education earned a baccalaureate degree by 2000 (Chen, 2005). This percentage compares to the 68% non-first generation students who had earned a baccalaureate degree in this timeframe. Students who are first in their families to attend college differ from their peers in a number of important ways. Although there are several models and research studies that explore the factors that lead to student attrition or graduation, they do not take into account these unique differences. Much of the research focuses on generally accepted success indicators (such as standardized test scores) for the majority student population, which predominantly consists of Caucasian students and rarely addresses at-risk student populations, such as first generation students (Pascarella, Pierson, Wolniak, & Terenzini, 2004). The purpose of this study was to determine whether a relationship exists between the background, cognitive, noncognitive variables (dropout proneness, predicted academic difficulty, educational stress, receptivity to institutional help) measured by the Noel-Levitz College Student Inventory (CSI) Form B, as well as participation in a first generation student support program and the dependent variables cumulative college grade point average, retention from first to second year, and baccalaureate degree attainment for first generation/first time freshman college students at a public research university in Texas. The quantitative predictive study examined five successive cohorts of students who enrolled at the university between 2002-2006. The study evaluated the entire sample of participants in 2010 to determine which factors significantly predicted academic success for the group of students. Results from the discriminant analysis indicated that dropout proneness, predicted academic difficulty, high school percentile, and gender were significantly related to students’ baccalaureate degree attainment. Discriminant analysis also showed that dropout proneness, educational stress, high school percentile, and predicted academic difficulty were significantly related to students’ retention status. Results from the multiple regression analysis indicated high school percentile, gender, and SAT/ACT standardized scores did significantly predict cumulative college GPA. Results from the multivariate analysis of variance showed that there was a statistically significant difference between Hispanic and Caucasian students on the combined noncognitive variables. Logistic regression analysis indicated that the noncognitive variables were not statistically significant better predictors than the cognitive variables for degree attainment within five years. High school percentile was the only significant predictor of degree attainment.