Table 2.

Study 1:Multi‐group analysis

Paths	Multi‐group Models						Chi‐square differences			Unconstrained model
	χ2	df	p	CFI	RMSEA [CI 90%]	SRMR	Δ χ2	Δdf	p	Low‐SES group			High‐SES group
										b	p	β	b	p	β
Community Id → Social Support (Baseline Model)	0.50	1	.482	1.00	.00 [.00, .08]	.006	–	–	–	.43	<.001	.45	.46	<.001	.48
Community Id → Loneliness	1.61	2	.448	1.00	.00 [.00, .07]	.008	1.11	1	.292	−.13	<.001	−.18	−.09	.007	.12
Social Support → Loneliness	0.74	2	.692	1.00	.00 [.00, .05]	.006	0.24	1	.624	−.15	<.001	−.20	−.13	<.001	−.18
Community Id → Well‐being	4.44	2	.108	.997	.00 [.00, .09]	.012	3.95	1	.047	.13	<.001	.14	.05	.170	.06
Social Support → Well‐being	0.51	2	.775	1.00	.00 [.00, .05]	.006	0.01	1	.903	.04	.259	.04	.05	.159	.05
Loneliness → Well‐being	0.55	2	.761	1.00	.00 [.00, .05]	.006	0.05	1	.821	−.51	<.001	−.41	−.53	<.001	−.46

Multi‐group Models describe the constrained models with the path reported in the first column to be equal across the two groups in addition to the path of the baseline model (in the first row). Coefficients in the unconstrained model (saturated model) describe the paths of low‐SES group and high‐SES group in a model with all the paths allowed to vary across the two SES groups.