Table 3.

Standardized coefficients of logistic regression to model regional predictors of Tree City USA participation. Larger standardized coefficients represent a larger effect relative to other variables in the model. Positive coefficients denote a positive effect on Tree City USA participation, and vice versa. Spatial eigenvectors used to reduce spatial autocorrelation in model residuals are not shown among the independent variables. Bold values are significant at P < 0.05; values in bold italics are significant at P < 0.01.

Region	Standardized coefficients						Model fit
Region	Income	Education	Owner occupancy	White	Age	Population	R²_McFadden
1-New England	*+1.54*	*+1.43*	*−3.48*	−0.78	+0.83	+ 1.23	0.21
2-Middle Atlantic	*+1.37*	*+1.01*	*−2.35*	−0.40	+0.05	+0.27	0.16
3-East North Central	*+0.79*	*+1.15*	*−1.54*	*+0.43*	+0.19	*+5.24*	0.15
4-West North Central	+0.40	+0.32	*−0.65*	*+0.74*	−0.02	*+6.70*	0.22
5-South Atlantic	*−0.80*	*+0.89*	*−1.03*	+0.31	+0.40	+6.39	0.21
6-East South Central	+0.61	−0.13	*−1.89*	*+0.91*	*+0.99*	*+7.45*	0.22
7-West South Central	+0.23	*+2.86*	*−1.34*	−0.04	+0.12	*+3.29*	0.26
8-Mountain	−0.33	*+0.91*	*−1.36*	*+3.23*	*−1.05*	*+2.72*	0.22
9-Pacific	+0.28	*+1.76*	*−1.40*	*+0.84*	*−0.72*	*+2.35*	0.15