Skip to main content

Advertisement

Table 2 Logistic Regressions Examining Demographic and Cannabis-Related Predictors of Quitting Status

From: Successful and unsuccessful cannabis quitters: Comparing group characteristics and quitting strategies

Predictor Odds ratio 95% confidence interval
   Lower Upper
Demographic    
   Age 1.01 0.98 1.35
   Gender 0.64 0.32 1.26
   Marriage (yes/no) 0.74 0.34 1.60
   University (yes/no) 0.34* 0.16 0.70
Cannabis-Related    
   Age/initiation 1.07 0.97 1.19
   Use (days/wk) 0.90 0.76 1.06
   Frequent exposure (yes/no) 6.04*** 2.21 16.55
   Previous quit attempts 1.01 0.98 1.04
   SDS 0.53* 0.32 0.88
   Formal Treatment (yes/no) 0.38* 0.16 0.92
  1. Demographics regression Nagelkerke R2 = .10, p <.001. Cannabis-related variables regression Nagelkerke R2 = .24, p <.001. Unsuccessful quitters coded as 0; successful quitters coded as 1.
  2. Gender: males coded as 0, females coded as 1; Marriage: not married coded as 0, married coded
  3. as 1; University: no coded as 0, yes coded as 1; Frequent exposure: no coded as 0, yes coded as 1.
  4. SDS = Severity of Dependence Scale. Interpretation of binary logistic regression depends on
  5. whether the predictor is categorical or continuous. The highest option is the reference point for categorical variables, whereas the lowest option is the reference point for continuous variables.
  6. Thus, odds ratios below 1 indicate successful quitters scored lower than unsuccessful quitters on a variable when it is continuous, and that successful quitters scored higher on a variable when it is categorical. Odds ratios above 1 indicate that successful quitters scored higher than unsuccessful quitters on a variable when it is continuous, and that successful quitters scored lower on a variable when it is categorical. *p < .05.