Table 2.
Meta-Regression Results of Andrews Principles on Drug Use Outcomes
| Bivariate Meta-Regression Models | Multiple Predictor Meta-Regression Models | |||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| k | B | S.E. | 95% CI | Beta | τ2 | k | B | S.E. | 95% CI | Beta | τ2 | |
| Risk Model | 231 | .03 | 231 | .03 | ||||||||
| Constant | .29 | .06 | .16, .42 | - | .28 | .07 | .14, .43 | - | ||||
| Risk | −.08 | .07 | −.21, .06 | −.06 | −.08 | .07 | −.22, .05 | −.06 | ||||
| Methods quality | −.06 | .04 | −.14, .02 | −.08 | ||||||||
| Publication status | .05 | .06 | −.06, .17 | .05 | ||||||||
| Needs* Model | 240 | .03 | 240 | .03 | ||||||||
| Constant | .19 | .03 | .15, .24 | - | .15 | .05 | .06, .24 | - | ||||
| Needs | .02 | .02 | −.02, .05 | .05 | .02 | .02 | −.02, .05 | .05 | ||||
| Methods quality | .−.06 | .04 | −.14, .01 | −.09 | ||||||||
| Publication status | .10 | .05 | −.00, .20 | .10 | ||||||||
| Responsivity Model | 124 | .02 | 124 | .02 | ||||||||
| Constant | .19 | .03 | .12, .26 | - | .18 | .07 | .05, .32 | - | ||||
| Responsivity | .02 | .02 | −.03, .06 | .05 | .02 | .02 | −.03, .06 | .05 | ||||
| Methods quality | −.01 | .06 | −.12, .10 | −.02 | ||||||||
| Publication status | .02 | .08 | −.13, .17 | .02 | ||||||||
| Appropriate Service Model | 59 | .02 | 59 | .02 | ||||||||
| Constant | .09 | .06 | −.03, .22 | - | −.01 | .12 | −.24, .23 | - | ||||
| Appropriate Service | .06 | .04 | −.02, .13 | .16 | .06 | .04 | −.02, .13 | .16 | ||||
| Methods quality | −.06 | .08 | −.22, .10 | −.08 | ||||||||
| Publication status | .15 | .13 | −.10, .40 | .12 | ||||||||
*
Number of criminogenic services for the E group minus the number of criminogenic services for the C group, then Winsorized at 7.