TABLE 2—

Regression Coefficients From a First-Stage Moderated Mediation Model of Depressive Symptoms on Incarceration History via Negative Police Encounters and Police Avoidance, Conditioned on Unemployment Among Black Men: Washington, DC, 2015–2016

	Coefficienta (SE or Bootstrap SE; 95% CI)	t
Outcome variable: depressive symptoms (R2 = 0.062; P < .001)
Incarceration history	−0.079 (0.058; −0.193, 0.034)	−1.376
Negative police encountersb	0.070 (0.020; 0.031, 0.109)	3.526
Police avoidanceb	0.070 (0.024; 0.024, 0.117)	2.976
Unemployment	−0.111 (0.050; −0.109, 0.087)	−0.223
Incarceration history × unemployment	0.108 (0.082; −0.052, 0.268)	1.323
Age	0.001 (0.003; −0.005, 0.006)	0.234
Marital status	−0.048 (0.040; −0.126, 0.031)	−1.189
Education	−0.033 (0.021; −0.074, 0.007)	−1.607
Conditional indirect effects via negative police encounters by unemploymentc
Employedb	0.049 (0.017; 0.018, 0.086)	. . .
Unemployedb	0.058 (0.021; 0.021, 0.103)	. . .
Conditional indirect effects via police avoidance by unemploymentd
Employedb	0.023 (0.011; 0.005, 0.046)	. . .
Unemployedb	0.047 (0.019; 0.013, 0.085)	. . .

Note. CI = confidence interval. The sample size was n = 843.

^aOrdinary least squares regression coefficient.

^bStatistically significant.

^cIndex of moderated mediation = 0.009 (SE = 0.013; 95% CI = −0.014, 0.037).

^dIndex of moderated mediation = 0.024 (SE = 0.014; 95% CI = 0.003, 0.055).