effects (Winer, 1971). To correct for the effects of multiple comparisons on Type I error rate, the family-wise α levels of post hoc contrasts were adjusted using the Holm-Bonferroni correction (Holm, 1979). For interactions of continuous (e.g., FTND) with categorical variables (e.g., valence), we calculated slopes for average change over the continuous variable, which are reported as point estimates (PE) and standard errors (SE). Point estimates are slopes that indicate change on a dependent measure for every 1-unit increase in the predictor, analogous to a regression beta statistic. Differences on demographic and baseline smoking behavior variables were evaluated by group using univariate ANOVAs.
