Preparation for Complex Sample Survey Data Analysis
This chapter guides the survey analyst through a sequence of steps that are important to properly prepare for the analysis of a complex sample survey data set. The steps include reviewing the weights and the sampling error calculation model for the survey data set, examining the amount of missing data for key variables, and preparing for the analysis of sample subclasses. In most cases, the data producer has access to far more detailed information about the complex sample design and personal information concerning study respondents than the data user. In everyday practice, it is always surprising to learn that an analyst who is struggling with the weighted analysis of survey data has never actually looked at the distribution of the weight variable. The chapter concludes with a short checklist to remind the reader of the critical preliminary steps.