Table 4.

Logistic Regression Analysis of an Arrest for a Crime (Other than Substance Use or Property-Related) in the Past Six Months on Demographics, Geographic Region, Stimulant Use, and Criminal Involvement (n=700)

	Coefficient (SE)	OR	95% CI
Demographics
Non-white	−.39 (.37)	.68	.33 – 1.40
Age*	−.03 (.02)	.97	.94 – 1.00
Male	.41 (.28)	1.51	.88 – 2.61
Single	−.12 (.27)	.89	.52 – 1.51
Years of Education*	−.12 (.06)	.89	.79 – 1.00
Employed	−.05 (.27)	.95	.56 – 1.63
Annual Income Category - $5,000 to $9,999a	−.18 (.31)	.83	.44 – 1.60
Annual Income Category – $10,000 or abovea	−.33 (.30)	.72	.40 – 1.30
Geographic Regionb
Arkansas*	.70 (.38)	2.01	.95 – 4.25
Kentucky*	.63 (.35)	1.88	.95 – 3.75
Last Six Month Pattern of Crack Usec
Less than Once a Month	−.57 (.46)	.56	.23 – 1.39
More than Once a Month, but Less than 3 Days per Week	.40 (.34)	1.49	.77 – 2.87
Three or More Days per Week**	.90 (.34)	2.47	1.28 – 4.77
Last Six Month Pattern of Powder Cocaine Usec
Less than Once a Month	−.15 (.34)	.86	.44 – 1.69
More than Once a Month, but Less than 3 Days per Week	−.15 (.33)	.86	.45 – 1.63
Three or More Days per Week	.15 (.37)	1.17	.56 – 2.42
Last Six Month Pattern of Methamphetamine Usec
Less than Once a Month	.20 (.35)	1.22	.61 – 2.43
More than Once a Month, but Less than 3 Days per Week	−.12 (.35)	.89	.45 – 1.75
Three or More Days per Week	.10 (.40)	1.11	.51 – 2.41
Last Six Month Pattern of Amphetamine Usec
Less than Once a Month	.55 (.51)	1.73	.64 – 4.70
More than Once a Month, but Less than 3 Days per Week*	.79 (.37)	2.21	1.06 – 4.59
Three or More Days per Week	−.12 (.67)	.89	.24 – 3.33
Criminal Involvement
Police Contact b/c Substance Use*	.63 (.32)	1.89	1.01 – 3.53
# of Convictions**	.03 (.01)	1.03	1.01 – 1.04

^aNOTES: Annual income less than $5,000 is the reference group;

^bOhio is the reference group;

^cNo stimulant use is the reference group; 2 Log likelihood = 515.29; Model χ² = 55.28, p<.001; Nagelkerke R² = .14;

^*p<.05,

^**p<.01 (one-tailed significance test)