Again, we begin with an intercept only model (Model 1). The variance component of the intercept is statistically significant (tau = .429, p < .01), indicating that perceived disorder, similar to fear of crime, varies significantly across the neighborhoods in the sample. Model 2 includes the demographic variables. The effect of gender is not statistically significant; females perceive similar amounts of disorder as males. In contrast, young and middle age (the reference category is the elderly), and education all significantly affect perceived disorder, suggesting that a mediating process for these variables exists. The young and middle aged, as well as educated people, perceive significantly more disorder. This is part of the reason that they are more fearful. Model 3 adds the two measures of vulnerability. Both effects on perceived disorder are not statistically significant. These findings indicate that self-described vulnerable people perceive the same amount of disorder, and hence perceived disorder is not a reason for them to be more fearful. These findings are consistent with the relatively stable coefficients for vulnerability across Models 4 and 5 in Table 1.Model 4 adds the two measures of victimization to the model. Both violent victimization (b = 1,052; p < .01) and personal property victimization (b = .492; p < .01) increased perceived disorder. Recall from Table 1, Models 5 and 6, that the effects of personal victimization on fear of crime become non-significant when the measure of perceived disorder is entered into themodel. These earlier findings, combined with those in Model 4 of Table 2, indicate that perceived disorder does in fact mediate the effects of victimization.Lastly, the Model 6 adds neighborhood-level disorder into the analyses. As expected, the coefficient is significantly positive (b=.219, p
đang được dịch, vui lòng đợi..