This study addresses limitations in pooled event history analysis (PEHA) by relaxing the assumption of policy homogeneity.
Context & Problem:
Standard PEHA assumes constant effects across policies, potentially masking important variations.
Methodology:
The authors employ Monte Carlo simulations to compare how different modeling strategies perform under varying levels of heterogeneity. ※ They systematically test these approaches with increasing variance in data.
Key Findings:※
Multilevel models with random coefficients emerge as the superior approach for handling policy differences, offering significantly better estimates than alternative methods.
This methodology provides a more nuanced understanding of how variables impact policies differently across contexts.
Practical Application:※
The paper demonstrates these techniques using a unique dataset tracking 29 anti-abortion policies over time. ※ Researchers can now better explore theoretical implications related to policy diffusion and design variation.
| Modeling Heterogeneity in Pooled Event History Analysis was authored by Rebecca Kreitzer and Frederick Boehmke. It was published by Sage in SPPQ in 2021. |
