Table 5. Negative Binomial regressions predicting recidivism from perceived and anticipated stigma.

	Perceived Stigma			Anticipated Stigma
	B	S. E.	Sig.	B	S.E.	Sig.
Arrest Diversity
Step 1
Race	.02	.30	.94	−.02	.31	.96
Type of Stigma	.18	.13	.16	−.10	.11	.36
Step 2
Race X Type of Stigma	−.35	.27	.20	−.41	.23	.08
Offense Diversity
Step 1
Race	.23	.26	.39	.26	.27	.34
Type of Stigma	.07	.11	.54	−.08	.09	.40
Step 2
Race X Type of Stigma	−.05	.21	.82	−.04	.19	.82

Note. Each type of stigma (perceived and anticipated) was used as a predictor for each outcome. Statistics for each type of stigma are reported in the appropriate column in this table.

Negative binomial regressions were conducted due to positive skewness in arrest and offense diversity variables.

N = 106-111 for arrest diversity and N = 105-110 for offense diversity.

¹Race was coded as 0-Caucasian, 1-African American.

^*p < .05