Handling Data with Three Types of Missing Values

Missing values present challenges in the analysis of data across many areas of research. Handling incomplete data incorrectly can lead to bias, over-confident intervals, and inaccurate inferences. One principled method of handling incomplete data is multiple imputation. This dissertation considers incomplete data in which values are missing for three qualitatively different reasons and applies a modified multiple imputation framework in the analysis of that data. The first major contribution of this dissertation is a derivation of the methodology for implementing multiple imputation in three stages. Also included is a discussion of extensions to estimating rates of missing information and ignorability in the presence of three types of missing values. Simulation studies accompany these sections to assess the performance of multiple imputation in three stages. Finally, this new methodology is applied to an insomnia treatment study with comparisons to other commonly used missing data methods