Aptitude, Motivation, and Self-regulation as Predictors of Achievement among Developmental College Students

Research & Teaching in Developmental Education, Fall 2003 by Ray, Marilyn, Garavalia, Linda, Murdock, Tamera

Abstract

The primary purpose of the present study was to examine the validity generalization of aptitude as a predictor of achievement for developmental college students. A second purpose was to investigate students' motivational beliefs and perceived self-regulated learning skills as predictors of achievement. Results indicated a weak correlation between aptitude and achievement, accounting for only .52% of the variance in achievement. As such, developmental educators may be using a criterion for admission that accounts for less than 1% of the variance in achievement. Together, aptitude, self-regulated learning, and motivational beliefs explained approximately 15.2% of the variance in achievement. Implications for education are discussed.

An estimated 650,000 students, or approximately one-third of the freshmen entering U.S. colleges and universities each year, are required to enroll in at least one developmental course in Reading, English, or Mathematics designed to prepare students to cope with college-level coursework (Boylan, Bonham, & Bliss, 1994). Measures of students' aptitude (ACT or SAT scores), their previous performance (high school grade point averages or class rank), or a combination of these factors traditionally have served as the basis for making decisions concerning admission and/or placement in developmental courses. Numerous studies have reported correlations between aptitude and academic performance in college coursework. Various researchers have studied the predictive validity of standardized measures of aptitude such as the ACT and SAT in content-area courses (Stallworth-Clark, Scott, & Nist, 1996), in undergraduate pre-service teachers' programs (Neal, Schaer, & Ley, 1990), and among female intercollegiate student-athletes (Petrie & Stoever, 1997). In these studies, SAT and ACT scores are predictive of course grades and semester grade-point-averages (GPAs). Harackiewicz, Barron, Tauer, and Elliot (2002) also reported that both aptitude and high school GPA were significant predictors of academic performance in college.

The predictive value of aptitude with regard to college achievement varies across studies. For example, Cote and Levine (2000) reported that IQ accounted for only 0 - 4% of the variance in academic achievement among Canadian university students. Likewise, Wolfe and Johnson (1995) found that prior performance (high school GPA) accounted for 19% of the variance in GPA among students enrolled in a college introductory psychology course, a motivational variable accounted for 9% of the variance while SAT scores accounted for only 5% of the variance. Other studies reported low to moderate correlations between measures of aptitude and academic performance. Britton and Tesser (1991) reported low correlations (r = .20) between college students' SAT scores and their cumulative GPA during their first two years of college. Further, Wentzel (1991) reported that student GPAs and SAT scores were only moderately correlated (r = .42) and that many of the students in the top 12% of their high school class did not score in the top 12% on the SAT. Moreover, Meeker, Fox, and Whitley (1994) tested the predictive value of 26 variables, including SAT scores, for college students' GPA in psychology. SAT scores (aptitude) were not significant predictors of GPA in psychology (the students' major), with Math SAT scores accounting for only 3.6% of the variance. In another study, measures of aptitude were found to be either unrelated or negatively related to achievement in college students, while motivation appeared to be more predictive of achievement (Cote & Levine, 2000).

With regard to developmental college students, a second potential limitation of prior research on the predictive validity of aptitude is the nature of the sample. Because these studies have been conducted with the general college student population, the generalizability of findings to developmental college student populations may be questionable. A number of studies have shown that at-risk or developmental college students exhibit personal characteristics that distinguish them from regular-admission college students. For example, Larose and Roy (1991) examined the role of prior academic performance and affective variables, such as fear of failure and test anxiety, in predicting the success of at-risk college students. Affective variables were more reliable predictors of academic success for at-risk students, but high school GPA more accurately predicted course grades for regularly-admitted students. Also, among freshmen enrolled in a developmental mathematics course, Higbee and Thomas (1999) reported that SAT scores and high school GPA were not significantly correlated with course grades, whereas students' beliefs about their ability (self-efficacy beliefs) were more closely aligned with their achievement.

Therefore, a potential explanation for low and inconsistent correlations between aptitude and achievement lies with the composition of the sample; perhaps the predictive value of aptitude varies significantly within the college student population, with higher predictive value for some subgroups, such as high-achieving students, and lower predictive value for others, such as at-risk or developmental students. As such, aptitude may not be the best predictor of developmental college student success. A purpose of this study is to determine the generalizability of aptitude as a predictor of academic achievement to developmental college students, a subgroup of the larger general college student population. A secondary purpose of the study was to investigate the predictive value of motivational beliefs and self-regulated learning variables as alternative criteria for admission decisions with developmental college students.