Skip to main content
. Author manuscript; available in PMC: 2020 Jan 1.
Published in final edited form as: Drug Alcohol Depend. 2018 Oct 23;194:59–68. doi: 10.1016/j.drugalcdep.2018.10.003

Table 4.

Multiple Regression Results Testing Predictive Validity of Three AUDIT Scoring Methods: Any 30 day binge drinking

Cutoff Sum Score MNLFA Score

Logistic Regression
Results
B (SE)
Ordinary Least Squares
Regression Results
B (SE)
Ordinary Least Squares
Regression Results
B (SE)
Intercept −1.41 (.15) −1.95 (.18) −.79 (.22)
Female −.90 (.07) −.27 (.06) −.17 (.08)
Black −.45 (.14) −.50 (.18) −.44 (.22)
Hispanic −.22 (.16) −.25 (.20) −.20 (.25)
White −.25 (.15) −.33 (.18) −.03 (.23)
CARE 1.02 (.12) 1.32 (.16) .73 (.16)
New Hope −.49 (.25) −.23 (.42) −.57 (.32)
STT .91 (.09) .05 (.18) .74 (.19)
Start Together .99 (.15) −.20 (.36) −.01 (.41)
Stride .07 (.23) −.26 (.38) −.27 (.42)
Success 1.07 (.29) −.15 (.69) .39 (.64)
Vista .63 (.18) −1.18 (2.42) −1.55 (.43)
AUDIT Score 3.86 (.29) .97 (.04) 4.98 (.26)
AUDIT Score x
CARE
−.41 (.44) −.58 (.06) −3.40 (.29)
AUDIT Score x
New Hope
−.71 (1.01) −.47 (.11) −3.40 (.42)
AUDIT Score x
STT
.43 (.59) .77 (.14) 1.02 (.58)
AUDIT Scorex
Start Together
- .96 (.33) 1.72 (1.30)
AUDITScorex
Stride
1.43 (1.09) .28 (.21) −.10 (.90)
AUDITScore x
Success
- .59 (.45) .75 (1.87)
AUDIT Score x
Vista
- .75 (.83) 2.83 (1.01)

Note. The BRIGHT study site was used as the comparison group in analyses because it had the largest sample size. Estimates for cells with dashes could not be obtained due to empirical underidentification. Bold parameters are significant at p<.05 but given that scores are on different metrics, values from the different regressions are not directly comparable. All effects are reported controlly for other reported variables in the model.