Table 3.
Primary outcomes by time and treatment condition: Random effects model estimates
| Variable | N, obs. | −2RLL | Denominator. df | F | p | Effect size, d |
|---|---|---|---|---|---|---|
| Self-reported percent days abstinent by month | ||||||
| Intercept | 120, 475 | 3100.139 | 215.58 | 517.86 | .00 | |
| Medication Condition (GAL vs PLA) | 215.58 | .45 | .50 | |||
| Behavioral Condition (CBT4CBT vs TAU) | 215.58 | 1.98 | .16 | |||
| Time | 351.07 | 152.04 | .00 | |||
| Contrast 1: Gal + CBT4CBT vs GAL+TAU and PLA+CBT4CBT by time | 351.07 | .03 | .86 | .03 | ||
| Contrast 2: Galantamine vs placebo by time | 351.07 | 5.27 | .02 | .34 | ||
| Contrast 3: CBT4CBT vs TAU by time | 351.07 | 4.22 | .04 | .30 | ||
| Number of cocaine-negative urine specimens submitted by month | ||||||
| Intercept | 120, 367 | 1495.842 | 267.52 | 21.58 | .00 | |
| Medication Condition (GAL vs PLA) | 267.52 | .01 | .91 | |||
| Behavioral Condition (CBT4CBT vs PLA) | 267.52 | .00 | 1.00 | |||
| Time | 286.14 | 32.17 | .00 | |||
| Contrast 1: Gal + CBT4CBT vs GAL+TAU and PLA+CBT4CBT by time | 286.14 | 5.04 | .03 | .75 | ||
| Contrast 2: Galantamine vs placebo by time | 286.14 | 3.46 | .06 | .43 | ||
| Contrast 3: CBT4CBT vs TAU by time | 286.14 | .11 | .74 | .03 | ||
Abbreviations: −2Rll=-2 restricted log likelihood. Gal=Galantamine, PLA=placebo, CBT4CBT=Computerized CBT, TAU=Treatment as Usual