Date on Master's Thesis/Doctoral Dissertation


Document Type

Master's Thesis

Degree Name



Bioinformatics and Biostatistics

Committee Chair

Smoot, Tanya M.

Author's Keywords

Public health


Surveys--Databases; Surveys--Statistical methods; Surveys--Data processing


Missing data is very common in survey research. However, currently few guidelines exist with regard to the diagnosis and remedy to missing data in survey research. The goal of this thesis was to investigate properties and effects of three selected missing data techniques (listwise deletion, hot deck imputation, and multiple imputation) via a simulation study, and apply the three methods to address the missing race problem in a real data set extracted from teh National Hospital Discharge Survey. The results of this study showed that multiple imputation and hot deck imputation procedures provided more reliable parameter estimates than did listwise deletion. A similar outcome was observed with respect to the standard errors of the parameter estimates, with the multiple imputation and hot deck imputation producing parameter estimates with smaller standard errors. Multiple imputation outperformed the hot deck imputation by using larger significant levels for variables with missing data and reflecting uncertainty with missing values. In summary, our study showed that employing an appropriate imputation technique to handling missing data in public use surveys is better than ignoring it.