Table 6.

Model coefficients for the moderated mediation models predicting symptoms of depression or anxiety

Antecedent	Consequent
	M (Self-concept)				Y (Depression)
	Path	Coeff	SE	p	Path	Coeff	SE	p
X (Migration status)	a	.035	.399	.931	c’	1.350	.384	< .001
M (Self-concept)	_	_	_	_	b1	-.450	.031	< .001
W (Age)	_	_	_	_	b2	4.767	1.794	.008
M × W	_	_	_	_	b3	-.138	.045	.003*
Constant	iM	38.569	.259	< .001	iY	26.155	1.268	< .001
Model	R2 = -.000				R2 = .229
	F (1, 1814) = .001, p = .931				F (4, 1811) = 134.439, p < .001

Antecedent	Consequent
	M (Self-concept)				Y (Anxiety)
	Path	Coeff	SE	p	Path	Coeff	SE	p
X (Migration status)	a	.040	.398	.922	c’	2.287	.421	< .001
M (Self-concept)	_	_	_	_	b1	-.242	.034	< .001
W (Age)	_	_	_	_	b2	6.045	1.963	.002
M × W	_	_	_	_	b3	-.169	.050	< .001**
Constant	iM	38.551	.260	.000	iY	22.679	1.389	< .001
Model	R2 = .000				R2 = .103
	F (1, 1820) = .001, p = .922				F (4, 1817) = 52.091, p < .001

Left column: Path a for predictor X (migration status) to mediator M (self-concept) calculated according to Hayes [56], R² and F-tests for differences to intercept-only model. Right column: Full model paths (c’, b₁, b₂, b₃) in moderated mediation models with symptom scores of depression or anxiety as consequent as illustrated in Fig. 1 Panel B), adjusted R² and F-test for differences to intercept-only model. See Results section for estimates on conditional effects. M, mediator. W, moderator. M × W, interaction effect. SE, standard error. Uncorrected p-values for estimates reflecting study hypotheses in bold, alpha-error corrected thresholds applied to indicate significance at ***p > .001, **p > .01, *p > .05