- Year 2024
- NSF Noyce Award # 1950292
- First Name Li
- Last Name Feng
- Registration Faculty/Administrator/Other
- Discipline STEM Education (general)
- Role Principal Investigator (PI)
- Presenters
Philip Ervin, Jieon Shim, Li Feng, and Cody Michalik, Texas State University; Steven Liu, University of California San Diego
Need
Historically marginalized groups are underrepresented in the STEM workforce (Bernard & Cooperdock, 2018), as evidenced by the significant STEM achievement gap between non-disadvantaged and economically disadvantaged student groups. On standardized tests, students of color and students from low-income families have lower test score averages. This disparity is because of a variety of socioeconomic factors both inside and outside the classroom. In an effort to recruit more STEM teachers in high-need school districts, the National Science Foundation created the Robert Noyce Scholarship Program to recruit undergraduate and graduate-level STEM majors to teach in these school districts (Porter et al., 2022). This project measures the Noyce Scholarship Program’s nationwide impact on math skills from the 3rd grade to the 8th grade from the year 2009 to 2019. We also examine if students who are economically disadvantaged or students from groups historically underrepresented in STEM benefit to a greater or lesser extent.
Research Questions
We are interested in answering two research questions related to the potential impact of the Noyce Scholarship Program on nearby public-school districts’ student achievement and test score gap between different racial groups. 1. Do high-need public school districts located within 10 or 25 miles of a Noyce institution perform better on standardized tests in math, reading, and language arts compared with high-need districts not located near Noyce institutions? 2. Are race-based and poverty-based achievement gaps smaller for high-need districts near Noyce institutions?
Approach
To identify school districts with Noyce scholars, we make the assumption that high-need school districts in close proximity to a Noyce institution will have a relationship with the institution and a pipeline to recruit Noyce scholars. To generate our geospatial data, we take the centroid of a school district’s boundary and find Noyce institutions within a 10, 25, and 50-mile radius. The high-need school districts with Noyce institutions within these radii are considered “treated.” High-need school districts outside these boundaries are considered controls. To determine school district characteristics and test score outcomes, we use an extensive dataset from the Stanford Education Data Archive (SEDA). This dataset provides different educational performance metrics across United States school districts from 2009 to 2019. The data captures district characteristics and student achievement outcomes by race, sex, and eligibility for free or reduced-price lunches.
Outcomes
In high-need school districts, the achievement gap between non-economically disadvantaged and economically disadvantaged students shrinks from the 3rd grade to the 8th grade mathematics. This gap is smaller among treated school districts. We find a similar trend in achievement gaps between white and black students, where the treated school districts have a smaller gap than districts located further from a Noyce institution. This is also true for the achievement gap between white and Hispanic students.
Broader Impacts
On a nationwide scale, we find empirical evidence that the Noyce Scholarship program addresses achievement gaps between economically disadvantaged and non-disadvantaged students, as well as between different racial and ethnic groups. The success of the Robert Noyce scholarship program underscores the importance of investing in teacher education and professional development. Even when considering only effects found at the middle-school level, the program demonstrates effectiveness in attracting high-quality teachers to high-need school districts.
