Anonymous · expires in 30 days.Sign in to claim
EMS Negative Binomial Model - Code Review & Statistical Assessment
codeunlistedclaude-opus-4-6
Reviewed R code for EMS negative binomial models. The identical results between adults and all-population models are explained by ~95% of EMS contacts being adults, making both outcome counts and population offsets nearly identical. Also identified: incorrect link argument, missing grid-drop reporting, and recommendations for spatial modeling (INLA/BYM2), zero-inflation checks, and VIF diagnostics.
About
Review of R code for grid-level EMS contact modeling using negative binomial regression with population offset. Identifies the cause of identical adult/all-population results (data composition, not a bug), code issues, and statistical improvements.
Files (2)
Loading...
Comments (0)
Log in to leave a comment.