First-year academic success: A prediction combining cognitive and psychosocial variables for Caucasian and African American students

Journal of College Student Development, Nov/Dec 1998 by Ting, Siu-Man Raymond, Robinson, Tracy L

Only three variables were found to be significant predictors for African American students. They were high school GPA, SAT-total scores and planning for work hours. Unlike the Fall GPAs, which found SAT-math scores as the predictor, high school GPA was found to be a significant predictor for African American males' Spring GPAs.

SAT-total scores and planning for work hours were found to account for 17.4% of the variance of the Spring GPAs for African American females, which were higher than those of the Fall variance (8%). SAT scores, instead of high school GPA were found to be significant predictors. Statistical results in the Spring are available upon requests to the first author.

Predicting Student Retention

A total variance of only 1.5% was found to explain students' retention after the Fall semester. This result was similar to previous studies (Fuertes et al., 1994; Tracey & Sedlacek, 1985, 1987). In addition, the small number of dropouts (n = 89) may be one of the reasons for such low variance. However, continuing studies for retention in the second year is needed when the enrollment information becomes available.

DISCUSSION

The prediction models in the current study explained a range of 8% to 29% of the variance for students' GPA in the first year. High school GPA was the most significant predictor for GPA in the first year, particularly for Caucasian students. Other studies reported similar findings (Houston, 1980; Stanley, 1971; Ting, 1997a). However, a variety of psychosocial variables were also found to be significantly correlated with students' GPA, such as parents' educational level, course load, planning for work hours, contact with university staff in the admission process, and intention to join extracurricular activities. Other studies findings also showed psychosocial variables as effective predictors (e.g., organized campus activities, involvement with peers and faculty, and institutional commitment (Milem & Berger,1997); community service and leadership experience (Ting,1998)).

As predicted, the current findings indicate that the multivariate models to predict academic performance across gender and race are more effective than a general model for the whole sample. Sedlacek (1996) also disputed the assumption that a single measure can predict equally well for all students. In the current study, both male and female students were found to be influenced by high school academic performance, self-concept, and future goals. On the one hand, these findings add to the current professional literature that revealed similar predictors for all students in their samples (Fuertes et al.,1994; Hood,1992; Tracey & Sedlacek,1984). On the other hand, convergence of race and gender in the current study revealed more psychosocial predictors which were not explored in these previous studies. For example, students' personal development is an important indicator that includes taking responsibility for one's behavior, improving leadership skills, coping with change, handling stress, time management, and selfdiscipline. These qualities seem necessary for students to adjust to college life successfully. Father's education was also significant in the prediction of Fall semester GPA for women as well as men. This might be explained by the role of fathers in influencing both sons and daughters to pursue college, given their own educational level and perhaps accompanying class privileges that allow them to financially support their families.