Ethnic and gender differences in science graduation at selective colleges with implications for admission policy and college choice

TitleEthnic and gender differences in science graduation at selective colleges with implications for admission policy and college choice
Publication TypeJournal Article
Year of Publication2004
AuthorsSmyth, F.L., McArdle JJ
JournalResearch in Higher Education
Volume45
Pagination353-381
KeywordsAcademic achievement, Academic preparation, ACT/SAT, Diversity Institute Literature Review, Gatekeeper courses, K-12, Mathematics, Minorities, Recruitment, Retention, Women
SummaryThis study seeks to explain the variance in science graduation rates based on gender and ethnicity as a function of students' pre-college academic preparation. The authors found that attending a selective college did not increase students' likelihood of graduation. Therefore, they propose that students of all demographic groups who have low to mid-range math SAT scores attend colleges where their peers are at a similar academic level to them (i.e., less-prestigious colleges).
Extended SummaryNational science organizations have expressed concern about the higher attrition rates of female and underrepresented minority (URM) students in science majors. The authors agree with Elliott et al. (1995), who believe that affirmative action processes at highly selective colleges result in minority students with lower math SAT scores opting out of the physical and biological sciences to graduate in other fields instead. However, they acknowledge that, as Bowen and Bok (1998) pointed out, minority students are more likely to graduate at more selective schools, even if they do not remain in the "science track." Based on a set of statistical models, the authors conclude that students with lower math SAT scores are actually better off attending less-selective colleges if they want to pursue a science degree. The authors matched demographic information with students' intended majors, and found that White females were the least likely of all groups to select science majors. Black students were twice as likely as White students to be interested in the field. In almost all ethnic categories, males were more likely than females to express interest. (Black female students were equally as likely to express interest as Black male students.) The authors also evaluated records of student persistence. Hispanic females were equally as likely to persist in their program as Hispanic males, but males were more likely to persist than females in all other ethnic categories. A 75-point increase in math SAT scores corresponded to a 50% greater chance of graduation in the sciences. The hypothesis that SAT score explained science persistence held for differences between White and URM students. However, gender has a residual effect that cannot be explained by SAT score alone. There also appear to be other variables, outside of the scope of the study, which explain Asian students' success relative to all other population groups. This study raises a number of questions about outside factors that the SAT and college courses have in common. The reliance of some science departments on multiple-choice testing, an assessment method that also appears on the SAT, may favor students whose learning style is more compatible with that type of testing. In addition, students' access to resources such as study groups and tutors, and their willingness to seek out assistance, may be related to their cultural background and gender. The assistance that students receive when studying math may contribute to their success on the SAT and, later, in college.
RecommendationsAlthough this paper does not offer instructional recommendations, the results indicate that additional assistance should probably be made available for students struggling with math concepts that they did not learn before entering college. The authors also recommend that affirmative action programs maintain strict standards regarding students' SAT scores.