After the survey was closed, responses were exported from Survey Monkey into SPSS 19.0 for analysis. Descriptive statistics were used to characterize the sample, and inferential statistics were used to identify relationships between variables. Initial analysis involved calculating Pearson's correlation coefficients among each of the measures of social integration and psychological distress. The correlations were followed by a principal components analysis (PCA) to determine the factor structure among the integration and distress variables. To determine the sample size needed for factor analysis, researchers have suggested using the ratio of participants to variables analyzed (see MacCallum, Widaman, Zhang, & Hong, 1999, for a review). For example, Gorsuch ticipants for each variable analyzed), Everitt (1975) suggested that a ratio 10:1 is adequate for estimating producing reliable factors. The ratio participants variables analyzed for the PCA conducted in the present study was 22:1, exceeding both Gorsuch's (1983) Everitt's (1975) suggested minimum requirements. multivariate analysis of variance then conducted examine mean differences between smokers and non smokers on the broad factors from the PCA.