. ResultsDescriptive data for potential moderator variables are reportedin Table 1. Descriptive statistics of the dependent variable (MASQ-GA) across assessment periods are reported in Table 2.3.1. Preliminary analysesNo significant differences existed between included versusexcluded participants on baseline MASQ-GA and moderatorvariables (p ¼ .99). Among participants included in the moderator analyses, there were no significant differences between ACT versusCBT (p ¼ .60) or between individuals with a principal diagnosis of panic disorder versus other disorders (p ¼ .66) on baseline MASQ- GA and moderator variables. Baseline MASQ-GA did not signifi- cantly correlate with any potential moderator variables (ps > .13).