Table 3.

Reliability of Change Coefficients and Standard Errors of Measurement

	Abstinence Self-Efficacy		Coping Strategies		Depression		Anxiety
No. Items (ni)	Coef.	SEM	Coef.	SEM	Coef.	SEM	Coef.	SEM
As administered*	.87	6.2	.90	0.10	.87	0.09	.86	0.15
1	.57	13.8	.22	0.53	.24	0.40	.40	0.44
2	.72	9.8	.37	0.37	.39	0.28	.57	0.31
3	.80	8.0	.47	0.30	.49	0.23	.67	0.25
4	.84	6.9	.54	0.26	.56	0.20	.73	0.22
5	.87	6.2	.59	0.24	.61	0.18	.77	0.20
6			.63	0.22	.66	0.16	.80	0.18
7			.67	0.20	.69	0.15	.82	0.17
8			.70	0.19	.72	0.14	.84	0.16
9			.72	0.18	.74	0.13	.86	0.15
10			.74	0.17	.76	0.13
11			.76	0.16	.78	0.12
12			.78	0.15	.79	0.12
13			.79	0.15	.81	0.11
14			.80	0.14	.82	0.11
15			.81	0.14	.83	0.10

Note. Coef.=reliability of change coefficient, SE_M=standard error of measurement for within-person change. Coefficients reflecting poor, fair, and good reliability of change are shown using roman, italic, and bold fonts, respectively. Coefficients reflect the amount of variance related to changes in scores over time that is attributable to reliable change. See text for details. For consistency in the table, SE_M values are always computed with the assumption that total scale scores are computed by averaging rather than summing items. For measures that are typically summed (i.e., depression, anxiety), the SE_M must be multiplied by number of items administered to properly compute the SE_M for summed scores.

^*As administered in the parent trial, the number of items was 5 (self-efficacy), 30 (coping strategies), 21 (depression), and 9 (anxiety).