Skip to main content
. Author manuscript; available in PMC: 2018 Aug 1.
Published in final edited form as: Psychol Addict Behav. 2017 Jun 19;31(5):636–646. doi: 10.1037/adb0000286

Table 3.

Results of Linear Mixed Models for Laboratory and Field data

n1 n2 Group Group x Time
df PE SE F p df PE SE F P
Lab Data
VP Bias (ms) 183 55 1,126 19.24 6.34 9.20 .003 1,125 10.02 12.14 0.68 .41
Stroop (ms) 93 55 1,37 51.39 32.94 2.43 .13 1,36 −28.73 49.35 0.34 .56
SABQ (0–4) 101 56 1,44 −0.07 0.15 0.25 .62 1,43 −0.02 0.21 0.01 .93
QSU (0–5) 101 56 1,44 0.13 0.24 0.28 .60 1,43 0.04 0.25 0.03 .86
CO (ppm) 105 57 1,47 −0.54 1.25 0.19 .67 1,46 0.02 1.50 0.00 .99
Cotinine (ng/mL) 101 57 1,43 −11.39 46.87 0.06 .81 1,42 51.25 50.58 0.32 .57
Field Data
VP Bias (ms) 481 53 1, 374 24.06 9.68 6.18 .01 1, 374 0.34 2.63 0.02 .90
Self-Reported Attention (1–7) 2249 56 1, 2081 −0.23 0.23 0.95 .33 1, 2081 0.02 0.02 1.12 .29
Advertisements (1–5) 2241 56 1, 2075 0.05 0.12 0.20 .65 1, 2075 0.00 0.12 0.00 .96
Cued Craving (1–7) 2276 56 1, 2108 −0.45 0.28 2.50 .11 1, 2108 0.02 0.03 0.44 .51
Craving (1–7) 2276 56 1, 2108 −0.39 0.28 2.00 .16 1, 2108 0.03 0.03 1.20 .27
Smoking (PDA) (1–5) 2225 56 1, 2057 0.02 0.24 0.01 .93 1, 2057 0.04 0.02 3.91 .04
No. Cigs (Diary) 738 54 1, 683 1.25 1.07 1.38 .24 1, 682 −0.01 0.05 0.05 .82

Note. n1 = number of level 1 units, i.e., assessments, visits, or days; n2 = number of subjects (level two units) who completed at least one assessment at visit 2 or visit 3 (lab data) and at least one PDA assessment (field data). Columns labeled Group show the results for the main effect of Group. The comparison category is CON (Negative parameter estimates indicate lower values for AR vs. CON). Columns labeled Group x Time shows the results for the Group x Visit (lab data) or Group x Day (field data) interaction term. Visit is a categorical variable with two levels (visit 2 vs. visit 3). All models include main effects for Group and Visit. The LMM for VP Bias also includes Picture Type (old vs. new) as an additional within-subject factor (lab data). In addition, the baseline (pre-intervention) measure for each dependent variable was included as a covariate. Data for analysis of VP Bias use all assessments with no more than one reported interruption (see text). Key: CO = expired carbon monoxide, QSU = Brief Questionnaire for Smoking Urges, SABQ = Subjective Attentional Bias Questionnaire, VP Bias = Visual Probe attentional bias score, PE = (unstandardized) parameter estimate; SE = standard error; F = F value from mixed model