Voluntary Association Memberships of American Adults: Evidence from National Sample Surveys

differentiation takes place. Knowledge of the association of the dimensions becomes important as a basis for further evaluations of the Shevky theory. In the reported factor analysis tests, the social rank, urbanization, and segregation factors were allowed to assume whatever relationships were necessary to best fit the predicted models. Factor correlations were computed from the original rotated matrices for all cities.26 As Table 5 indicates, these correlations show a rather consistent pattern. The low correlations, ranging from -.143 in Akron to .224 in Atlanta, between social rank and urbanization factors, indicate that they are almost independent of each other. The segregation factor, however, is related to both social rank and to urbanization. Correlations of segregation with social rank are consistently negative for the separate cities and range from -.015 in Atlanta to -.752 in Providence. Segregation is positively correlated with urbanization, with values varying from effectively zero in Portland to .344 in Akron.