Skip Navigation
HomeSHS OverviewMethodsTrainingAnalysisDisseminationResources


Data Analysis

After the field work and analysis, data will need to be entered into a database-management system. This may include questionnaires, data-collection forms and laboratory results. It is important to plan data management before beginning the data collection.Careful planning can prevent errors in data analysis. 

Statistical Analysis: Once data are entered into a data-management system, statistical analysis can be conducted in order to make inferences about research questions. Analysis, using statistical software, can range from descriptive statistics to multivariate analysis, depending on the objectives of the study.

  • Descriptive Statistics: Descriptive statistics include frequencies, percentages, means, medians, interquartile ranges, average smoker density and others. For example, SHS concentrations can be compared across cities, business types and locations within a building. Boxplots and other types of figures are often effectively used to describe and compare distributions of SHS exposure levels.
  • Multivariate Analysis: Multivariable analysis considers multiple variables at one time. For example, the association between smoking policy and air nicotine concentrations in bars and nightclubs can be measured after controlling for location, establishment size, occupancy and ventilation patterns. Regression models often are used to adjust for multiple predictors of outcome such as air nicotine concentrations.


This project is funded by the Bloomberg Initiative to Reduce Tobacco Use and the Flight Attendant Medical Research Institute (FAMRI), developed in consultation with Roswell Park Cancer Institute and the University of Southern California, Institute for Global Health.