Table 3.

Random effects regression on cross-cultural training – panel data (1995 and 2005)

	Outcome variables
	Cross-cultural training (Percent of staff trained)	Cross-cultural training (Number of hours)
	Tobit (SE)	Tobit (SE)
Independent variables
Supervisors
Supervisor's cultural sensitivity	0.78 (0.23) ***	0.10 (0.03) ***
Public revenue x cultural sensitivity	-2.27 (12.16)	-1.01 (1.70)
Directors
Director's tenure	0.04 (0.38)	0.01 (0.05)
Director's license	9.78 (5.78) *	0.96 (0.81)
Director in community	2.10 (0.92) **	0.22 (0.12) *
Director's education	-5.57 (2.55) **	-1.03 (0.35) ***
Control variables
Client racial/ethnic diversity in unit
High - Latino/a	---	---
Medium - Latino/a	---	---
High - Blacka	3.38 (7.25)	-0.30 (1.00)
Medium - Blacka	6.36 (7.50)	0.60 (1.04)
Organizational factors
% public revenue	19.18 (8.84) **	0.58 (1.22)
State license	-12.42 (11.23)	-0.02 (1.51)
Accreditation (JCAHO)	5.91 (9.00)	0.10 (1.25)
Unit size (number of clients)	1.92 (2.90)	-0.76 (0.41)*
Number of services	---	---
Staff resources	19.11 (16.17)	1.71 (2.26)
Methadone	1.19 (7.43)	-0.96 (1.03)
Urban unit	0.43 (1.47)	0.20 (0.20)
Ownership
For profitb	-21.80 (10.83) **	-4.13 (1.52) ***
Non-for profitb	1.40 (7.70)	-1.73 (1.05)
Affiliation
Unit in hospitalc	-11.16 (10.88)	-3.10 (1.51) **
Unit in mental health facilityc	-9.55 (8.50)	-2.42 (1.18) **
2005 waved	36.84 (7.18) ***	-1.51 (1.00)
Constant	105.0 (44.30) **	18.56 (6.10) ***
Sigma_u	25.45 (10.12) **	0.01 (2.50)
Sigma_e	82.81 (4.34) ***	13.36 (0.30) ***
Observations	1184	1184

Note. Unstandarized parameter estimates, with standard errors in parentheses from two tailed tests.

^aUnits with low percentage (< 10%) of Black clients is the referent.

^bPublic is the referent.

^cFree standing unit is the referent.

^d1995 wave is the referent.

^*p < .05

^**p < .01

^***p < .001