TABLE 3.

Results of a Stepwise Logistic Regression Predicting Incidence of Fraud

Variables	Step 1				Step 2				Step 3
	β	SE	Wald χ2	Exp(B)	β	SE	Wald χ2	Exp(B)	β	SE	Wald χ2	Exp(B)
Age	−.02	01	5.95*	.978	−0.2	01	410*	.979	−.01	01	1.11	.989
Gender	−.16	15	1.06	.855	−.21	06	1.67	.815	−.20	16	1.59	.816
Minority status	.08	21	15	1.083	.03	21	02	1.1030	−.05	21	.05	.953
Education	.06	03	4.62*	1.059	.08	03	7.71**	1.086	.09	03	8.67**	1.093
Marital status	.12	17	47	1.123	.03	17	04	1.033	−.07	18	.16	.933
2002 CES–D					.11	04	7.38**	1.118	.09	04	4.54*	1.094
2002 ADL					.02	13	02	1.019	.01	13	.00	1.006
2002 IADL					−.16	30	29	.850	−.18	30	.36	.834
2002 Self-rated health					−.01	09	02	.988	−.07	09	57	.935
2002 Income					.00	00	68	1.000	.00	00	.53	1.000
2002 TICS					−.01	02	28	.991	−.01	02	.20	.993
2002 Disease burden					.04	08	30	1.044	.03	08	.12	1.028
2008 Financial satisfaction									−.27	07	13.67***	.764
2008 Affection									.01	03	.14	1.013
2008 Confirmation									.06	05	1.75	1.066
2008 Status									−.07	03	4.99*	.932
Constant	−2.19	78	80	.11	−2.47	85	8.41**	.08	−2.15	88	6.02*	.117

Gender is coded as 0 = male, 1 = female Minority status is coded as 0 = White, 1 = non-White Marital status is coded as 0 = partnered, 1 = not partnered

^*p < .05;

^**p < .01;

^***p < .001.