College persistence and student attitudes toward financial success

College Student Journal, June, 2005 by Karen Leppel

What can be done to prevent attrition by students with ill-suited majors? Schools need to provide information to assist students in selecting fields that fit their talents and personalities. Students should be encouraged to explore different courses to see what appeals to them. The career counseling office should have assessment tools that compile students' personal attributes and identify compatible educational and career goals. Then, if the appropriate match is a lucrative field, the student will be even happier. But if it's not, and the student is one who places much emphasis on financial success, the counselor needs to explain that the student is likely to be more satisfied in the long run in a well-suited career. Furthermore, people are less likely to be successful in jobs they dislike. Ultimately, if students receive and follow appropriate major and career advice, the students and the institution will be better off.

Appendix

Since academic performance and persistence are simultaneously determined, performance would not be independent of a persistence equation error term, and a two-step process is appropriate. In the first step, a performance equation was estimated by least squares regression. Then, the value of performance predicted from that equation was used as an explanatory variable in the persistence equation, which was estimated using logit analysis. Estimation was performed using NCES weights, which compensate for potential bias due to ineligibility and nonresponse among the students in the BPS sample.

Tinot (1993, pp.77) suggested that college persistence might vary with gender. Females, he wrote, "are more likely than males to face external pressures which constrain their educational participation." Because the impact of other independent variables on persistence may be different for male and female students, the current study estimates separate pairs of performance and persistence equations for males and females. Using the data described above, the performance and persistence equations were estimated.

The performance equation, which was estimated using least squares regression, was

GPA = [a.sub.0] [a.sub.1] ASIAN [a.sub.2] BLACK [a.sub.3] HISPANIC [a.sub.4] AGE [a.sub.1] HIGHACAD [a.sub.6] DRIVE [a.sub.7] INTEG [a.sub.8] HOURS [a.sub.9] MARRIED [a.sub.10] KIDS.

ASIAN was 1 for Asians and 0 for non-Asians. BLACK was 1 for blacks and 0 for non-blacks. HISPANIC was 1 for Hispanics and 0 otherwise. MARRIED was 1 for married students and 0 for unmarried students. KIDS was 1 for students with children and 0 for students without children. HIGHACAD was 1 for students who perceive themselves as above average academic ability, and 0 otherwise. DRIVE was 1 for students who perceive themselves as above average in drive to achieve, and 0 otherwise. AGE, INTEG, and HOURS were as described previously.

The persistence equation was estimated using logit analysis and the CATMOD procedure of SAS. The equation was:

ln[[p.sub.j]/[p.sub.3]] = [b.sub.0] [b.sub.1] ASIAN [b.sub.2] BLACK [b.sub.3] HISPANIC [b.sub.4] AGE [b.sub.5] DRIVE [b.sub.6] INTEG [b.sub.7] FAMINC [b.sub.8] HOURS [b.sub.9] MARRIED [b.sub.10] KIDS [b.sub.11] PARCOLL [b.sub.12] FINATT [b.sub.13] PREDGPA,