Table 2.

Regression coefficients for multi-level logistic model predicting candidate preference from candidate attractiveness and participant age, functional limitations, and education

Variable	Beta	SE	CI Lower	CI Upper	z	p	Effect size*	SD
Intercept	−0.114	0.281	−0.665	0.437	0.41	0.685
Main Effect of Attractiveness	0.448	0.079	0.293	0.603	5.64	< .001	3.79	4.14
Main Effect of Subject Age	0.195	0.503	−0.791	1.181	0.39	0.699	0.20	0.50
Main effect of Functional Limitations	−0.003	0.019	−0.040	0.034	0.16	0.876	−0.02	3.75
Education	0.082	0.055	−0.026	0.190	1.48	0.139	0.27	1.59
Attractiveness x Subject Age	−0.484	0.116	−0.712	−0.256	4.17	< .001	−6.61	6.69
Attractiveness x Functional Limitations	−0.012	0.007	−0.026	0.002	1.94	0.052	−1.35	55.19
Subject Age x Functional Limitations	−0.005	0.033	−0.070	0.060	0.17	0.868	−0.10	10.24
Three-way Interaction	0.035	0.009	0.017	0.052	3.84	< .001	6.75	95.88

Note. Total df = 1855; election df = 53; subject df = 35;

^*Effect sizes are calculated as standardized regression coefficients using the method described in Snijders and Bosker (1999), applying standardized regression coefficients to multilevel models.