Bottia, Martha C., Stearns, Elizabeth, Mickelson, Roslyn A., Moller, Stephanie, & Valentino, Lauren
Growing the Roots of STEM Majors: Female Math and Science High School Faculty and the Participation of Students in STEM
University of North Carolina at Charlotte
What is the role of the demographics of high school faculty, more specifically the proportion of female math and science teachers, on college students’ decisions to declare and/or major in STEM?
Journal Name or Institutional Affiliation:
Economics of Education Review
Vol. 45 Pp. 14-27
- There is a positive and significant relationship between proportion female math and science teachers and students 'chances of declaring physical sciences, engineering and/or mathematics as a major.
- Men are more likely to declare a PSEM (physical sciences, engineering, and mathematics) than women, although women are more likely to declare biological sciences as majors.
- Attending a school with a higher proportion of female math and science teachers is related to a significant increase in the chances of declaring, physical sciences, engineering, or mathematics as a major, particularly for female students.
- The influence of proportion of female math and science teachers is even stronger for students’ odds of graduating with a STEM major than for a students’ chances of declaring a STEM major.
- In the case of high skilled women, their chance of graduating with a biology and PSEM major increase 44% when they move from attending a school that has a proportion of female math and science teachers of .54 to one that has a proportion of female math and science teachers of .72 (1 s.d. above the mean distribution of female math.
- The proportion White of a school has a negative significant relationship for students odds of declaring a STEM major and graduating in STEM.
- For almost all subsamples, proportion of free and reduced lunch students has no significant association with declaring a STEM major or graduating in STEM.
- The results suggest that although the proportion of female math and science teachers at a school has no impact on male students, it has a powerful effect on female students’ likelihood of declaring and graduating with a STEM degree, and effects are largest for female students with the highest math skills.
Journal Article Empirical Research
College, Occupational Outcomes, Outcomes, Racial Composition, SES Composition, STEM Major, Teachers
Method of Analysis:
Multilevel Models, Multinomial Logistic Regression
2004 North Carolina Public School Graduates
Unit of Analysis:
College, School, Student
- The North Carolina Roots of STEM Success dataset. This dataset contains longitudinal information on the academic performance and scholastic experiences of all 2004 North Carolina public school graduates who also matriculated at one of the 16 campuses of the University of North Carolina system. It contains data from grades 7 to 12 for the cohort that graduated in 2004. They used a sample of approximately 12,550 students (7250 women and 5300 men) that came from about 270 different high schools.
- The dependent variables were whether a student 1) declared or 2) graduated with a STEM major. The authors used multinomial dependent variables, where 0 indicates no declaration/graduation of a STEM major, 1 represents declaration/graduation with a major in biological sciences, and 2 indicates declaration/graduation with a major in any other STEM discipline other than biological sciences (such as physical sciences, engineering and mathematics).
- The key independent variable was the proportion of science and math teachers who were women among the math and science faculty of the high school from which each student graduated. The authors focus specifically on the math and science teachers because these are the ones who could be available to serve as possible role models, mentors, or who could most directly challenge the chilly climates towards STEM that might otherwise exist at the high school. In addition, because students are required to take algebra1, algebra2, and biology, there is certainty that there was at least some interaction between these teachers and students.
- Their models include student demographic and family characteristics such as race/ethnicity, gender, and SES (defined as whether student received a need-based Pell grant in college, and is (or not) a first-generation college student in his or her family).
- In all models, the authors also control for variables correlated with probability of declaring and/or graduating with a STEM degree and their primary independent variables. These controls include racial composition of the school (proportion of white students at school); proportion of female students in the school; proportion of students in advanced college preparatory courses; proportion of students receiving free/reduced lunch; and school locale (urban, suburban, or rural).
- In addition, the authors include a set of variables that aim to capture important aspects of the high school that could be associated with students’ interest in STEM, including measures of teachers’ experience, teachers’ education (advanced degrees and licensure), and teacher turnover (percent of teachers employed in a school when the students are in a grade who are no longer employed in the same school when students are in following grade).
Links high school composition to majoring in STEM.