Abstract
The literature on high school exit exams has found both positive and negative effects of these high stake exams on high school graduation rates. To this point, the literature has not taken into account the embedded nature of school districts within state education systems. We employ a Bayesian Hierarchical SLX model to account for the hierarchical nature of education data in the United States. Our approach also allows us to account for spatial spillovers that influence graduation rates across districts and states. Using school district and state-level data for 45 states and 8194 school districts in the U.S. in 2015, we generally find no statistically significant effect of state exit exams on high school graduation rates. Random effect coefficients, however, point towards high school exit exams being negatively associated with graduation rates in a handful of states.
Original language | English |
---|---|
Pages (from-to) | 189-206 |
Number of pages | 18 |
Journal | Review of Regional Studies |
Volume | 50 |
Issue number | 2 |
State | Published - 2020 |
Keywords
- Bayesian statistics
- Hierarchical modelling
- Spatial dependence
- State exit exams