Applied Missing Data Analysis in the Health Sciences

Preface

With a strong practical emphasis on health science applications, this book describes statistical methods and models for the analysis of data with missing values. We attempt to write so that researchers with experience in applied data analysis, but less technical knowledge than a statistician, should be able to understand and implement most of the methods described. For those with a stronger background in statistics, we provide more technical details as to not detract from the flow of rest of the chapter. We have also tried to choose examples that are relevant to most health science researchers who work in a variety of disciplines.

In all fields of study, missing data are a common problem since, for any data collection process, there are so many things that could go wrong that missing values are all too likely. Thus, when attempts are made to answer the scientific questions of interest, researchers ask the all-too-common question: what do we do with the missing data?

The statistical literature to answer this question is well developed, but overly technical and complicated for researchers who are not experts in statistics and methodology. Therefore, researchers may recognize the existence of missing data, but fail to respond for two reasons: first, they may not understand the consequences of ignoring missing data and how it can impact the validity of their results; second, there is a lack of understanding of the statistical methods for missing data and how to apply them ...