2010 - The Role of Peers and Grades in Determining Major Persistence in the Sciences

Attribution: Ost, Ben
Researchers: Ben Ost
University Affiliation: Cornell University
Email: bw292@cornell.edu
Research Question:
This paper examines the determinants of entering and then persisting in physical and life science majors. Also, it investigates the impact of one's peers on major persistence.
Published: Yes
Journal Name or Institutional Affiliation: Economics of Education Review
Journal Entry: Vol. 29, Pp. 923-934
Year: 2010
Findings:
  1. In both the physical and life sciences, students are “pulled away” by their high grades in non-science courses and “pushed out” by their low grades in their major field.
  2. In the physical sciences, females are found to be more responsive to grades than males, consistent with psychological theories of stereotype vulnerability.
  3. However, females do not appear to be more responsive to grades than males in life sciences. This could be because do not suffer from “minority” status within the life sciences compared to the physical sciences.
  4. For Hispanic and Black students, the raw gap in the likelihood of declaring a physical or life science major disappears after controlling for preparation/ability, suggesting that differences in intended majors between Black/Hispanic students and White students is entirely due to differential preparation between these groups.
  5. Hispanic students are found to have much lower persistence rates in the life sciences than white students.
  6. Consistent with the notion that stronger students tend to enter physical sciences, a one standard deviation increase in SAT score is associated with an 8.3% point increase in the likelihood of intending to major in the physical sciences and a 2.8% point decrease in the likelihood of intending to major in the life sciences.
  7. Entering college with calculus credit strongly influences the probability of intending a physical science major.
  8. When controlling for performance factors, females still are more likely to drop out of the physical sciences.
  9. While both males and females benefit from exposure to higher quality peers in physical sciences, the effect is more than twice as large for women as compared to men.
  10. Controlling for other variables, females are less likely to intend to major in physical sciences.
  11. Students who are at most risk of failing to persist are also most influenced by their peers in physical sciences.
Scholarship Types: Journal Article Reporting Empirical ResearchKeywords: College Major ChoiceEducational EconomicsGPAHuman CapitalPeer EffectsPersistenceScienceRegions: N/AMethodologies: QuantitativeResearch Designs: Administrative DataAnalysis Methods: Descriptive StatisticsMultinomial Logistic RegressionPropensity Score Matching Sampling Frame:College Students
Sampling Types: Non-Random - PurposiveAnalysis Units: ClassroomStudentData Types: Quantitative-Longitudinal
Data Description:
  • The data came from a large elite research university. This data encompasses the entire universe of this university, including the complete transcript of courses and grades for every entering student from 1997 to 2003.These transcripts also include unique course identifiers; thus, the author can exactly identify each student’s peers in every course they take.
  • The final cohort is followed until 2008 when most (but not all) students have either graduated or withdrawn. This transcript data is matched to admissions data such as SAT scores, class rank and demographic information. The analytic sample was 15,508 students.
  • IVs- Gender, race, number of incoming college courses, incoming calculus credit, high school percentile, SAT score, GPA in 1st year life sciences courses, GPA in 1st year physical sciences courses, class size in core physical sciences courses, class size in core life sciences courses. They also control for cohort fixed effect and course fixed effect.
  • DVs- Intending to major in physical science, life science, or non-science major which was declared major at entry. Also, persistence in physical science and persistence in life science which was measured by intending to major in a particular subject and then graduating in that subject. The author also an included a measure of peer persistence.
  • The author also investigates the existence of peer effects in one’s courses on persistence by using a propensity score analysis. The propensity score is a linear combination of pre-college characteristics. Using the propensity score for each student, the author calculated the average propensity score by class.
Theoretical Framework:
Relevance:Factors Related to STEM Readiness, STEM Persistence and Retention
Archives: K-16 STEM Abstracts