Item Details

abstract

The American Community Survey (ACS) is one of the most vital public sources for demographic and socioeconomic characteristics of communities in the U.S. administered by the U.S. Census Bureau every year. The ACS provides communities with timely information by publishing 1-year and 5-year ACS time-period estimates. However, the time-period estimates may not adequately represent the yearly trend because they take an average over the multi-year period and are subject to uncertainty due to the sampling design. Many epidemiology and public health studies use the ACSestimates as explanatory variables in the models. However, using the ACS estimates as explanatory variables ignores the uncertainty in the time-period estimate and may disguise temporal volatility due to averaging over the time period. Our study will propose a model that accounts for the uncertainty in the ACS time-period estimates and disentangle the temporal misalignment in the multi-year time-period estimates. In order to achieve this goal, we will use a Bayesian hierarchical model that accounts for spatial or spatio-temporal dependency in data. We present the approach by modeling the prevalence of frequent mental distress in North Carolina from 2014 to 2018.

subject

autocorrelation

autoregressive models

Bayesian hierarchical models

change of support

spatial random effects

time-period estimates

contributor

Kwon, Jihyeon (author)

Hepler, Staci (committee chair)

Erhardt, Robert (committee member)

Evans, Ciaran (committee member)

date

2022-05-24T08:35:50Z (accessioned)

2022-05-24T08:35:50Z (available)

2022 (issued)

degree

Mathematics and Statistics (discipline)

identifier

http://hdl.handle.net/10339/100717 (uri)

language

en (iso)

publisher

Wake Forest University

title

A SPATIO-TEMPORAL HIERARCHICAL MODEL TO ACCOUNT FOR UNCERTAINTY IN AMERICAN COMMUNITY SURVEY EXPLANATORY VARIABLES

type

Thesis