Table 4.
Bayesian logistic regressions on cluster membership on abstinence by clustering algorithm. The Bayes facor of 5.88 indicates moderate evidence favoring the superiority of the 2-cluster solution over the 3-cluster solution.
| k-means Clustering | BNP+ | ||||||
|---|---|---|---|---|---|---|---|
| OR* | HPD# | Posterior probability < 0 | OR | HPD | Posterior probability < 0 | ||
| C>P & vs. P>C ^ | 0.58 | 0.28 – 0.91 | .983 | 0.62 | 0.28 – 0.98 | 0.961 | |
| C>P vs. New Cluster | 0.61 | 0.24 – 1.08 | 0.069 | ||||
| P>C vs. New Cluster | 1.02 | 0.47 – 1.74 | 0.442 | ||||
| DIC @ | 369.37 | 371.89 | |||||
| Bayes Factor (k-means vs. BNP models) | 5.88 | ||||||
Bayesian Non-Parametric;
Odds Ratio;
Highest Posterior Density;
larger brain responses to cigarette-related cues than non-nicotine-related pleasant stimuli;
larger brain responses to non-nicotine-related pleasant stimuli than cigarette-related cues;
Deviance Information Criterion