Logistic Regression: A Self-learning Text, Third Edition (Statistics in the Health Sciences)

Objectives Upon completing this chapter, the learner should be able to:

1. Given a binary logistic model involving two or more
   exposures, describe or illustrate how to carry out a
   modeling strategy to determine a “best” model.


2. Given a fitted binary logistic model involving a large
   number of exposure and/or covariates (potential
   confounders or effect modifiers), describe or illustrate
   how to conduct screening to reduce the number of
   variables to be considered in your initial multivariate
   model.


3. Explain by illustration when it is questionable to screen
   covariates using statistical testing for a crude
   association with the outcome variable.


4. Given a binary logistic model involving several
   exposures and/or covariates, describe and/or illustrate
   how to assess collinearity and how to proceed if a
   collinearity problem is identified.


5. Given a binary logistic model involving several
   exposure variables and/or covariates, describe and/or
   illustrate how to determine whether there are any
   influential observations and how to proceed with the
   analysis if influential observations are found.


6. Given a binary logistic model involving several
   exposure variables and/or covariates, describe and/or
   illustrate how to consider (or possibly correct for)
   multiple testing when carrying out a modeling strategy
   to determine a “best” model.

Objectives 243