Beyond Admission: Understanding Pre-College Variables and the Success of At-Risk Students

Journal of College Admission, Summer 2007 by Mattson, Christopher Erik

Abstract

This study examined pre-college variables from an admission-office perspective and the ability of these variables to predict college grade point average (GPA) for students specially admitted into an academic support program for at-risk students. The research was conducted at a private, highly-selective, research university in the southwest United States. The primary determining factors for this special admission program are lower-than-average high school GPA and/or standardized test scores. Pre-college variables that most significantly predicted college GPA were high school GPA, gender of student, and leadership experience prior to applying. Scholastic Aptitude Test (SAT) scores failed to predict success as measured by college GPA.

Beyond Admission

Seventy-five percent of students who drop out of college do so during their first two years, and 57 percent of students leave their first college without graduating (Tinto, 1993). First semester grades (McGrath & Braunstein, 1997} and first year grades (Gifford. Briceno-Perriott, & Mianzo, 2006; Reason, 2003) are significantly linked with retention. Because these grades act as a quantifiable predictor of retention and because grades are associated with academic success, this study focuses on the predicting of first-semester and first-year GPA of at-risk students.

The use of high school GPA and rank is widely accepted as a positive predictor of academic success (Astin, 1997; Hoffman & Lowitzki, 2005; Schwartz & Washington, 2002; Stricker, Rock & Burton, 1996; Ting, 1998; Wolfe & Johnson, 1995). Standardized test scores have also been found as a predictor, although they have been questioned in recent years (Astin, 1997; Lawlor, S., Richman, S. & Richman, C.L., 1997; Naumann, Bandalos & Gutkin, 2003; Reason, 2001; Stricker, et, al, 1996). Student involvement (Astin, 1984) and a variety of emotional and social variables (Boulter, 2002; DeBerard, Spielmans & Julka, 2004; House, Keely & Hurst, 1996; McGrath & Braunstein, 1997; Ridgell& Lounsbury, 2004; Spitzer, 2000; Wolfe & Johnson, 1995) have also been recently demonstrated as possible predictors of success.

Changes in demographics have altered studies on academic success and retention (Reason, 2001). Much of the research has been based on the traditional view of white, 18- to 22-yearold, full-time students, even though the number of students of color in higher education increased 61 percent between 1984 and 1994 (Pascarella & Terenzini, 1998). These changes have created the necessity for research that understands the new demographics of higher education.

Literature Review

A review of the literature showed that research related to the predicting of academic success determined by GPA in college has been productive in recent years, but not yet progressive. Significant findings have been made using academic-related variables (Lawlor, et. al, 1997; Reason, 2001, 2003; Stricker, et. al, 1996), non-academic variables (DeBerard, et. al, 2004; Naumann, et. al, 2003; Spitzer, 2000) and a combination of both to predict academic success (McGrath & Braunstein, 1997; Ridgell & Lounsbury, 2004; Schwartz& Washington, 2002; Ting, 1998; Wolfe & Johnson, 1995). Despite these findings, and numerous recommendations, the usage of pre-college variables remains very much the same (Astin, 1975, 1984, 1997; Atkinson, 2001; Cooper, 1999; Fleming & Garcia, 1998; Lawlor, et. al, 1997; Organ, 2001; Pascarella & Terenzini, 1998; Speyer, 2004; Tarn & Sukhatme, 2004; Tinto, 1993).

The non-academic related factors primarily researched included emotional health, social health and physical health. A study by DeBerard, Spielmans and Julka (2004) examined 10 variables that encompassed academic factors, social-support, coping methods, and health status. Their findings indicated a correlation with their variables for 56 percent of the variance of first year GPA. Only low high school GPA, however, could be significantly associated with attrition.

Positive predictors of GPA in a study of 355 full-time undergraduates by Spitzer (2000) were academic efficacy, self-regulation and social support. Naumann, Bandalos and Gutkin (2003) examined first-generation college students. In their questionnaire study of 155 students they were able to identify self-régulât ion as a positive predictor. They also found that ACT scores were positively correlated with the GPA of first-generation students.

Test scores remained under attack. The validity (Speyer, 2004; Ting, 1998), usage (Atkinson, 2001; Cooper, 1999; Reason, 2001; Tarn & Sukhatme, 2004) and fairness (Fleming & Garcia, 1998) of standardized test scores, such as the SAT and ACT were being questioned. Cooper (1999) and Reason (2001) examined ways in which adjustments could be made to the scores to accommodate for diversity and differences between high schools.

Reason, who proposed a merit-index score (2001), significantly predicted the academic achievement of white and African-American students with an ACT-based merit-index. Cooper (1999) addressed the "strivers" approach introduced by the Educational Testing Service (ETS). He defined strivers as applicants who exceed the scores of individuals from similar backgrounds by 200 or more points. Comparable to the merit-index approach, the problem with this strategy is that it failed to live up to its goal of offering more opportunities to underrepresented populations. The students who lose spots to strivers were often minority students attending more affluent schools.