Date of Award
Fall 2020
Document Type
Open Access Dissertation
Department
Biological Sciences
First Advisor
Alexander C McLain
Abstract
Analyzing population representative datasets for local estimation and predictions over time is important for monitoring related public health issues, however, there are many statistical challenges associated with such analyses. Mixed effect models are one of the common options which can incorporate time and spatial effect in the model and related inference is well established.
In the first part of this dissertation, to estimate area-level prevalence using individuallevel data, small area estimation (SAE) with post-stratified mixed effect models were used where sampling weights were also incorporated into it. However, if poststratification which requires more computation effort can improve estimation accuracy is not clear given the complicated modelling framework. Therefore, comparing the mean squared prediction errors (MSPE) to evaluate the predictive ability of post-stratification is of interest. In this study, various bootstrap methods were also implemented to calculate confidence intervals for post-stratified estimates, and investigating and comparing the performances of different bootstrap methods is another aim of this study. Under different model complexity situations, we are able to identify the best-performed bootstrap methods in the simulation study.
The second part of the dissertation involves analyses and predictions of disease prevalence using a penalized B-spline model. A unique feature of the data is that the sampling standard errors (SSEs) coming with the prevalence estimates need to be incorporated into the model. In previous studies, the uncertainty of the SSE is ignored which could influence the reliability of the estimation. In this study, we incorporate the uncertainty of the SSE and proposed an approximated likelihood function for fast computation. The performances of the proposed method were compared with some standard approaches in a simulation study.
Rights
© 2020, Yuan Hong
Recommended Citation
Hong, Y.(2020). Incorporation and Measurement of Uncertainty in Clustered and Spatial Data. (Doctoral dissertation). Retrieved from https://scholarcommons.sc.edu/etd/6188