2016 - Examining STEM Bachelor’s Degree Completion for Students with Differing Propensities at College Entry

Attribution: Wolniak, Gregory C.
Researchers: Gregory C. Wolniak
University Affiliation: New York University
Email: gwolniak@nyu.edu
Research Question:
1) What aspects of students' demographic, socioeconomic, and academic backgrounds influence selecting a STEM major during the first year of college? 2) To what extent do students' first year propensities toward a STEM education moderate the influence that first year financial aid and declaring a STEM major, as well as academic performance and integration in the academic and social environments have on STEM degree completion?
Published: Yes
Journal Name or Institutional Affiliation: Journal of College Student Retention
Journal Entry: Vol. 18(3) 287–309
Year: 2016
Findings:
  1. In terms of high school academics, a consistent pattern was revealed in which grade point averages, ACT scores, level and years of mathematics coursework, and level of science coursework were all significantly higher on average among 6-year STEM bachelor’s degree completers than among the full sample of 6-year bachelor’s degree completers or among the full sample of college enrollees.
  2. Students who completed a STEM degree within 6 years reported slightly lower cumulative GPA’s than the average among graduates in all fields but higher cumulative GPA’s than the average among all first time 2003 to 2004 college enrollees.
  3. Female students had significantly lower odds of declaring a STEM major in their first year of college.
  4. Net of all other variables in the model, Black students relative to non-underrepresented minority students (i.e., White and Asian students) were more likely to declare a STEM major early in their college experience, as were students having English as their second language.
  5. While the descriptive results show that Black students were underrepresented among STEM bachelor’s degree completers relative to the full sample of college enrollees, the predictive models show that, net of high school academic and other pre-college variables, Black students were more likely to select a STEM major in the first year of college than were traditional racial or ethnic groups.
  6. The majority of STEM graduates recognized and acted on their interest in a STEM education very early in their college experience.
  7. For students who enter college with a collection of background characteristics typically incongruent with majoring in a STEM field, it is critically important to foster interest in STEM fields during the first year of college.
  8. In terms of high school academics, every measure included in the model proved to be a significant predictor of declaring a STEM major. The highest level of mathematics taken exerted the largest effect on students’ likelihoods of selecting a STEM major, while the years of mathematics taken during high school had a smaller and opposite influence. Given the cumulative nature of mathematics content, it may be that simply counting years of math in high school, net of the highest level completed, captures those students who required additional work in lower levels of the math curriculum.
  9. Results highlight the lasting influence of high school academic achievement, the importance of declaring a STEM major early in a student’s college career, and disparate effects of academic performance and levels of social and academic integration.
  10. For some students (those who entered college in the middle STEM propensity category), receiving relatively low levels of grant aid (versus no grant aid) was a deterrent for STEM degree completion.
  11. Students’ initial propensities toward STEM careers appear to moderate the influence of grant aid on STEM degree completion, though the relationships are relatively weak.
  12. The findings suggest that students who initially may not be oriented toward STEM fields upon entering college may particularly benefit from receiving information on and encouragement toward STEM programs of study during their first year of college.
Scholarship Types: Journal Article Reporting Empirical ResearchKeywords: Bachelors Degree CompletionCollege Major ChoicePersistenceSTEMSTEM MajorRegions: NationalMethodologies: QuantitativeResearch Designs: Secondary Survey DataAnalysis Methods: Descriptive StatisticsLogistic Regression Sampling Frame:College Students
Sampling Types: Nationally RepresentativeAnalysis Units: CollegeStudentData Types: Quantitative-Longitudinal
Data Description:
  • The study incorporates elements from Social Cognitive Theory, economic and sociological perspectives, and models related to college student involvement and departure. Social Cognitive Theory provides a framework for examining career development and educational decision making by identifying that careers develop according to individual and contextual factors that exert influence at different points in time.
  • Beginning Postsecondary Students Longitudinal Study (BPS:04/09) on a nationally representative sample of first-time college enrollees during the 2003 to 2004 academic year.
  • The final analytic sample included approximately 7,330 students who were first time enrollees in a 4-year college or university during the 2003 to 2004 academic year. After accounting for missing data, analyses were based on weighted data representing roughly 1,396,520 students in the national population of 2003 to 2004 post-secondary enrollees.
  • The study’s primary outcome of interest was baccalaureate degree completion in a STEM field by the end of the sixth year after college entry, among all 2003 to 2004 first-time, 4-year enrollees. The comparison group included students who completed a bachelor’s degree in a non-STEM field as well as those who did not complete a bachelor’s degree 6 years after initially entering college.
  • Three types of measures comprised the study’s independent variables, including (a) whether a student declared a STEM major within his or her first year of college or not, (b) level and type of financial aid received during the first year of college, including total amounts of grants and loans received, and (c) college academics and integration. Academic integration was an eight-item composite measure based on mean scores from four 2004 and four 2006 survey items that indicated how often a student did the following: (a) participated in study groups, (b) had social contact with faculty, (c) met with an academic advisor, or (d) talked with faculty about academic matters outside of class.
  • Social integration was similarly derived as a six-item composite measure of three 2004 and three 2006 survey items indicating how often the student: (a) attended fine arts activities, (b) participated in intramural or varsity sports, and (c) participated in school clubs.
  • The measure of college GPA utilized was based on a 4.0 scale and coded into seven categories.
  • Student-level control variables consisted of demographic and socioeconomic measures, including gender, race or ethnicity, English as a second language, first generation U.S. citizen or immigrant status, parental education attainment based on either parent having attained at least a bachelor’s degree, and household income (parents’ income for financially dependent students or personal income for financially independent students). Addition covariates included measures of high school academic performance including GPA (4.0 scale), composite ACT or converted SAT score, highest level of mathematics completed, and the number of years of mathematics and science completed.
  • Institutional-level controls included institutional selectivity based on the level of selectivity of the first institution the respondent attended during 2003 to 2004, coded into three categories: (a) very selective, (b) moderately selective, and (c) minimally selective, open admission, or missing; Control (public 4-year; private not-for-profit 4-year; private for-profit 4-year); tuition and fees, coded into quintiles based on sample distribution; and percentage of underrepresented minority students (i.e., Black, non-Hispanic, American Indian or Alaskan, and Hispanic) enrolled at the institution in 2003 to 2004, also coded into quintiles based on the sample distribution.
Theoretical Framework:
Relevance:STEM Persistence and Retention
Archives: K-16 STEM Abstracts