Table 2. Multilevel models with wave and demographic variables predicting the log odds of social interactions.

	No One		Partner		Family		Friends
N observations	15,201		10,295		15,201		15,201
N participants	188		124		188		188
	Estimate	SE	Estimate	SE	Estimate	SE	Estimate	SE
Fixed effects
Intercept	0.62*	0.29	-1.29**	0.50	-1.53***	0.41	-2.68***	0.37
Wave 2	0.12	0.12	0.37*	0.16	0.09	0.15	-0.79***	0.20
Wave 3	0.26**	0.09	0.32**	0.12	0.02	0.12	-0.72***	0.15
Wave 4	0.36***	0.09	0.27*	0.13	-0.17	0.13	-0.57***	0.16
Control variables
Age	-0.26**	0.10	0.56***	0.15	-0.34*	0.14	0.07	0.13
Gender (male)	-0.01	0.16	0.20	0.21	-0.79***	0.22	-0.03	0.20
Race (white)	0.15	0.20	-0.38	0.33	0.01	0.28	0.60*	0.25
Living with spouse or partner	-1.09***	0.18	--	--	0.00	0.25	-0.99***	0.23
College	0.08	0.17	-0.39	0.25	-0.26	0.24	0.03	0.21
Working	-0.05	0.17	-0.30	0.24	-0.30	0.25	-0.60**	0.22
Income 1 ($30,000–$59,999)	-0.31	0.22	0.51	0.37	0.14	0.31	0.54	0.27
Income 2 ($60,000–$99,999)	-0.39	0.26	0.32	0.38	0.72*	0.37	0.20	0.33
Income 3 (> = $100,000)	-0.21	0.25	0.21	0.36	0.71*	0.36	0.71*	0.33
Random effects
Intercept Var: wave:person	0.23		0.29		0.35		0.52
Intercept Var: person	0.79		0.98		1.56		0.90
AIC	18289.09		11800.48		11820.23		7085.72
BIC	18403.53		11901.83		11934.67		7200.16
R2 conditional	0.29		0.34		0.40		0.36
R2 marginal	0.07		0.09		0.05		0.08

Note.

* p < .05

** p < .01

*** p < .001. Wald tests were conducted only on fixed effects. Var = variance.