Project Activities
The researchers will conduct this project through a partnership between RAND and the Kentucky Center for Statistics (KYSTATS). It will involve secondary data analysis of Kentucky's longitudinally linked education data ranging from preschool through postsecondary education and the workforce. The research team will use these data to describe how students' educational experiences vary with geography and conduct statistical tests of factors that contribute to different experiences. They will also conduct interviews and focus groups with school and district staff as well as students to understand the rural high school context and student experiences. They will focus on malleable factors and explore potential for policy changes that can improve student experiences. In particular, the research team will focus on measuring and understanding the reasons for (1) gaps in educational attainment between rural and nonrural students, (2) geographic differences in the distribution of school staff, (3) access to advanced high school courses and postsecondary education, and (4) the implications of college costs and financial aid for students' educational attainment in rural areas. Throughout the project, the researchers will engage with the other partners and advisory board members in Kentucky, as well as local stakeholders, to ensure the analyses appropriately consider the local context and to help disseminate the findings.
Structured Abstract
Setting
The study will take place in Kentucky including both rural and non-rural settings.
Sample
The primary sample consists of the cohorts of Kentucky public school students expected to complete high school between 2009 and 2027. Among these cohorts, the researchers will use the full population of Kentucky students, which may consist of approximately half a million students.
Some of the malleable factors for this study include geography (for example different rural distinctions), school support personnel factors (for example, how many, what type), student access to advanced high school coursework, access to postsecondary education (for example, distance to campus), the cost of postsecondary education (including financial aid), and the availability of online postsecondary education options.
Research design and methods
The research team will conduct exploratory analyses using administrative education data from Kentucky and mixed methods. For the administrative data, they will use several quantitative analyses. They will first summarize patterns in student experiences and achievement using descriptive approaches. They will use multivariate regression models to explore how differences in student outcomes are related to student, school, and community characteristics. They will also use Oaxaca decomposition methods to measure how our malleable factors of interest contribute to observed education gaps. Finally, they will conduct some analyses using quasi-experimental methods including regression discontinuity designs and difference-in-differences. In addition to the quantitative work with the administrative data, the researchers will also conduct interviews with school and district staff in Kentucky and focus groups with current and recent high school students. They will use a comparative case study methodology to identify the similarities and difference within and between rural and nonrural schools.
Control condition
Because the analyses are all exploratory, there is no pure treatment-control contrast. In some parts of the research, the researchers may consider students in nonrural parts of Kentucky the comparison group for students in rural areas, they may consider students with limited access to certain factors (such as online education and school staff) the baseline group relative to students with more access.
Key measures
The primary outcome measures are students' ACT scores, grade point averages (GPAs), advanced course enrollment, course grades, college enrollment, the types of postsecondary institutions students attend, financial aid received, persistence in college, degree attainment, time to degree, income, and employment rates. The primary predictor of interest is rurality, which the researchers will measure using NCES locale codes and census measures of population density and geography. For the moderators, the main measures are access to school personnel (defined using full-time equivalent and person allocations across schools), access to advanced high school courses (defined according to high school offerings and the availability of online options), access to postsecondary education (defined based on distance and availability of online options), cost of postsecondary education and financial aid received, internet access (defined using the FCC benchmark), and student characteristics (such as demographics, socioeconomic status, and baseline achievement).
Data analytic strategy
The researchers will use methods such as multivariate regressions, t-tests, Hausman tests, and likelihood ratio tests. They will measure how malleable factors contribute to educational gaps using multivariate regression and Oaxaca decomposition methods. They will explore potential solutions to gaps between rural and nonrural areas using multivariate regressions and regression discontinuity designs. They will code our qualitative data thematically and conduct analysis by research question, theme, participant type or role, and school type.
People and institutions involved
IES program contact(s)
Project contributors
Products and publications
In addition to providing preliminary evidence the possible interaction between geography (ruralness) and education outcomes, the project team will also produce journal articles, policy briefs, and presentations for research and non-research audiences, including local stakeholders.
Publications:
ERIC Citations: Find available citations in ERIC for this award here.
Questions about this project?
To answer additional questions about this project or provide feedback, please contact the program officer.
