Statistical and methodological issues in the analysis of complex sample survey data: Practical guidance for trauma researchers

Authors

  • Brady T. West

    Corresponding author
    1. Center for Statistical Consultation and Research at the University of Michigan–Ann Arbor, Ann Arbor, MI
    • 3550-C Rackham Building, University of Michigan, Ann Arbor, MI 48109
    Search for more papers by this author

  • The author wishes to thank the organizers of the 2007 Conference on Innovations in Trauma Research Methods (CITRM) for the opportunity to present the material in this paper at the conference, and two anonymous reviewers for detailed and thoughtful comments on earlier drafts.

Abstract

Standard methods for the analysis of survey data assume that the data arise from a simple random sample of the target population. In practice, analysts of survey data sets collected from nationally representative probability samples often pay little attention to important properties of the survey data. Standard statistical software procedures do not allow analysts to take these properties of survey data into account. A failure to use more specialized procedures designed for survey data analysis can impact both simple descriptive statistics and estimation of parameters in multivariate models. In this article, the author provides trauma researchers with a practical introduction to specialized methods that have been developed for the analysis of complex sample survey data.

Ancillary