Abstract:
|
The purpose of this presentation is to discuss statistical methods for model-building using complex survey data. The balanced repeated replications method with user-supplied weights for variance estimation will be addressed and illustrated using the Tobacco Use Supplement to the Current Population Survey data with 160 replicate weights. The methods will be demonstrated via a recent study conducted to estimate the prevalence of unassisted smoking cessation (i.e., quitting without pharmacological aids or behavioral interventions), and identify the most common smoking cessation methods used by US adult smokers while trying to quit smoking between 2007 and 2011. Among successful quitters, factors associated with using pharmacological methods and quitting unassisted were examined via a survey logistic regression. Considered socio-demographic factors included age, gender, race/ethnicity, highest level of education, marital status and employment status. Additional factors included visiting a doctor twelve months prior to quitting and survey mode. Analyses were performed using SASĀ®9.4.
|