Before examining our hypotheses, we conducted analyses to determine whether our physiological and in-session ratings measures were sensitive to the valence of our slide stimuli at baseline, prior to treatment regimen randomization. In-session ratings (craving, negative mood, and positive mood) were self-report measures collected after every third slide, and slide valence was coded as the slide immediately preceding the in-session ratings. We conducted mixed models analyses separately for each measure at baseline (Session 1), including all subjects, with slide valence, regimen, and their interactions terms with slide valence as the fixed effects, and subject as the random effect.
