Which variables, among those that are statistically significant at an alpha of 0.05, have the
greatest and least relative importance on the fitted model?  why? thanks a lot.

parameter	df	estimate	error	chi-square	pr>chisq	estimate
intercept	1	0.111	0.0304	25	<.0001
pred1	1	0.97	0.0385	634	<.0001	0.2
pred2	1	0.000072	3.39E-06	449	<.0001	-0.29
pred3	1	-0.56	0.115	24	<.0001	-0.04
pred4	1	-0.004	0.003	1.6	0.2	-0.01