Table 3.
Logistic regression model predicting cannabis vape-pen use (including near significant/significant variables, omitting cannabis use frequency).
Without cannabis use |
|||
---|---|---|---|
O.R. | 95% C. I. | p-Value | |
Positive expectancies | 1.04 | 1.01–1.06 | 0.012 |
Drinks per drinking day | 12.49 | 1.48–105.65 | 0.020 |
Cannabis peer injunctive norms | 3.75 | 1.79–7.86 | < 0.001 |
Notes: Demographics (race and gender) and other variables (sensation seeking, lack of premeditation, social interaction anxiety, and cannabis injunctive norms-father, past week cigarette smoking) were included in initial regression models but were not near significant or significant predictors and thus were omitted from the final model. Cannabis use frequency was the only significant statistical predictor of vape-pen use when included in model. OR = odds ratio; C.I. = confidence intervals.