Table 7.

Fixed effects from multi-level models concerning selected item: ‘Race/ethnicity affects interactions between police and young people locally’.

	Model 4a		Model 4b		Model 4c
Variable	Estimate [95% CI]	p (FDR)	Estimate [95% CI]	p (FDR)	Estimate [95% CI]	p (FDR)
Time	0.17 [0.12, 0.23]	<.001	0.17 [0.12, 0.23]	<.001	0.18 [0.11, 0.26]	<.001
Age	–	–	0.07 [0.02, 0.11]	.020	0.08 [0.03, 0.14]	.007
Non-Male (vs. Male)	–	–	−0.02 [−0.14, 0.09]	.924	−0.03 [−0.16, 0.10]	.890
Black (vs. White)	–	–	0.38 [0.26, 0.51]	<.001	0.42 [0.28, 0.56]	<.001
Hispanic (vs. White)	–	–	−0.05 [−0.19, 0.08]	.632	−0.06 [−0.21, 0.09]	.632
Other Race (vs. White)	–	–	0.25 [0.07, 0.42]	.016	0.25 [0.05, 0.45]	.033
Time*Age	–	–	–	–	−0.04 [−0.10, 0.01]	.222
Time*Non-Male (vs. Male)	–	–	–	–	0.02 [−0.12, 0.15]	.966
Time*Black (vs. White)	–	–	–	–	−0.08 [−0.22, 0.07]	.449
Time*Hispanic (vs. White)	–	–	–	–	0.02 [−0.14, 0.17]	.966
Time*Other Race (vs. White)	–	–	–	–	0.00 [−0.20, 0.19]	.997

Note: CI, confidence interval; p(FDR), adjusted p-value using false discovery rate. Each model also contained random intercepts for recruit.