Auraria campus
Brady Lamson’s MS in Statistics Project Presentation
Wednesday, December 2 · 12:30 pm–2 pm
at Student Commons 4017

Title: Small Area Estimation with the Fay-Herriot Model Abstract: The Fay-Herriot (FH) model has been in use since 1979 to improve the reliability of small area estimates (SAE). It balances direct survey estimates with predictions of small area means based on a linear regression model, using the known sampling variances to determine the degree of adjustment. The FH model traditionally assumes independence of area-level random effects, which is an assumption that may be violated in cases where the response exhibits spatial autocorrelation. This project uses a hierarchical Bayesian framework, an intrinsic autoregressive (ICAR) model for the spatial random effects, and auxiliary information taken from remote sensing data in the NASS Crop Data Layer to adjust American Community Survey (ACS) estimates of county-level housing vacancy rates in Iowa. Model estimates were evaluated against housing vacancy rates from the decennial census. Results demonstrate a reduction in mean squared error (MSE) and improvements to estimate precision through the use of spatial random effects, though the additional remote sensing covariate provided limited improvement.
Source: cu-denver · updated automatically. Times are Denver time. Doors, prices and lineups change — the ticket link is the truth.