Modeling Strategy Guidelines
We begin this chapter by giving the rationale for having a strategy to determine a “best” model. Focus is on a logistic model containing a single dichotomous exposure variable which adjusts for potential confounding and potential interaction effects of covariates considered for control. A strategy is recommended which has three stages: (1) variable specification, (2) interaction assessment, and (3) confounding assessment followed by consideration of precision. The initial model has to be “hierarchically well formulated,” a term to be defined and illustrated. Given an initial model, we recommend a strategy involving a “hierarchical backward elimination procedure” for removing variables. In carrying out this strategy, statistical testing is allowed for assessing interaction terms but is not allowed for assessing confounding. Further description of interaction and confounding assessment is given in the next chapter (Chapter 7).
KeywordsInitial Model Modeling Strategy Product Term Elimination Procedure Surgical Wound Infection
Unable to display preview. Download preview PDF.