Table 4.
Trimmed multiple predictor HLM model of the effects of predictors (t0) on smoking outcome (t1)
| Fixed Effect | Coefficient | SE | OR | 95% CI | T-ratio | Approx. df | p-value |
|---|---|---|---|---|---|---|---|
| Temptation vs. Abstinent | |||||||
| Intercepta | −0.06 | 0.24 | 0.94 | [0.58-1.54] | −0.22 | 105 | 0.820 |
| Baseline Nicotine Dependencea,b | 0.04 | 0.02 | 1.04 | [1.01-1.08] | 2.54 | 105 | 0.013 |
| Recent smoking (prior to t0, 1=yes, 0=no) | −0.19 | 0.16 | 0.82 | [0.60-1.12] | −1.22 | 3206 | 0.223 |
| Time between reports (t0 to t1) | 0.02 | 0.02 | 1.02 | [0.96-1.08] | 0.79 | 3206 | 0.428 |
| Time since quit day (0=week 1; 1=week 2-3) | −0.77 | 0.12 | 0.46 | [0.36-0.58] | −6.59 | 3206 | <.001 |
| Agitationa | 0.20 | 0.09 | 1.22 | [1.02-1.48] | 2.18 | 106 | 0.031 |
| Distress | −0.05 | 0.10 | 0.95 | [0.78-1.14] | −0.52 | 3206 | 0.600 |
| Urgea | 0.48 | 0.08 | 1.60 | [1.35-1.92] | 5.42 | 106 | <.001 |
| Alcohola | −0.46 | 0.52 | 0.62 | [0.22-1.74] | −0.90 | 106 | 0.367 |
| Smoking vs. Temptation | |||||||
| Intercepta | −0.94 | 0.23 | 0.39 | [0.24-0.62] | −4.04 | 105 | <.001 |
| Baseline Nicotine Dependencea,b | −0.004 | 0.02 | 1.00 | [0.96-1.03] | −0.21 | 105 | 0.834 |
| Recent smoking (prior to t0, 1=yes, 0=no) | 1.24 | 0.14 | 3.46 | [2.58-4.63] | 8.37 | 3206 | <.001 |
| Time between reports (t0 to t1) | 0.14 | 0.03 | 1.15 | [1.08-1.23] | 4.18 | 3206 | <.001 |
| Time since quit day (0=week 1; 1=week 2-3) | 0.56 | 0.13 | 1.75 | [1.35-2.28] | 4.22 | 3206 | <.001 |
| Agitationa | −0.28 | 0.10 | 0.75 | [0.60-0.92] | −2.70 | 106 | 0.008 |
| Distress | 0.26 | 0.10 | 1.30 | [1.06-1.60] | 2.58 | 3206 | 0.010 |
| Urgea | −0.14 | 0.10 | 0.87 | [0.72-1.06] | −1.40 | 106 | 0.164 |
| Alcohola | 1.31 | 0.49 | 3.71 | [1.40-9.89] | 2.66 | 106 | 0.009 |
Multiple predictor HLM model including level one and level two covariates. Non-significant variables were pruned from the model. SE=Standard Error. OR=Odds Ratio. 95%CI=confidence interval.
Random coefficient. All other predictors were treated as fixed to facilitate model parsimony and convergence.
Baseline nicotine dependence measured by the total score on the WISDM-68. Similar results are obtained using the FTND total score (temptation vs. abstinent: coefficient=0.22, SE=0.09, OR=1.24, 95% CI [1.02-1.50], t(105)=2.28, p=.02; smoking vs. temptation: coefficient=0.02, SE=0.09, OR=1.02, 95% CI [0.85-1.22], t(105)=0.24, p=.80).