. Author manuscript; available in PMC: 2017 Oct 1.

Published in final edited form as: J Consult Clin Psychol. 2016 Jul 25;84(10):907–912. doi: 10.1037/ccp0000126

Table 3.

Multiple Linear Regression Analysis and Partial Correlations on Percentage of Cocaine-Negative Urine Toxicology Screens

	B	SE B	β	sr²	p
Baseline frequency	−2.15	0.31	−0.47	−0.46	.00
Attendance	0.29	0.08	0.23	0.23	.00
Study protocol	0.82	2.43	0.02	0.02	.74
Homework	0.21	0.08	0.17	0.17	.01
(constant)	51.86	9.87
Total R²	0.34
Total Adjusted R²	0.33
Total R	0.59
Total F(6, n-7)	19.80
p	.00
n	157

Partial Correlations between Homework Adherence _a and Cocaine-Negative Urine Toxicology Screens_b, Controlling for Attendance_c and Baseline Cocaine Use Frequency_d

Study	r_ab.cd	p	df	n

A	−.02	.89	32	36
B	.21	.11	56	60
C	.32	.03	44	48
D	.39	.24	9	12
A,B,D	.16	.10	105	109
All	.21	.01	153	157

Note. Bold text denotes significance at p < .05. Baseline frequency = days of self-reported cocaine use in 28 days prior to trial. Attendance = percentage of sessions attended. Homework = percentage of homework adherence.