Two-way multivariate analyses of variance (MANOVAs) were used separately for the reference, absolute and relative θ power measures. The respective θ power measures at the MESOIs were the multiple dependent variables, with group (NACs vs. LTAAs) as the between-subjects variable and stimulus (target vs. rare non-target vs. standard non-target) as the repeated-measures variable. Planned repeated contrasts were used to investigate significant stimulus effects. Additional one-way MANOVAs and multivariate analyses of covariance (MANCOVAs) were used to investigate the group effect for each level of stimulus. For the MANCOVA analyses, the P3b amplitude at electrode site Pz was used as the covariate in the analysis. The choice of this ERP amplitude component was based on previous findings on this data that showed that it was the best discriminator of the two groups (Andrew and Fein, in preparation), (Fein and Chang, 2006). The reference θ power measures at MESOIs were also used together with the P3b amplitude as covariates for the MANCOVA on relative θ power measures. For all MANOVAs (MANCOVAs), separate analyses were applied for the power measures derived from ST-based and STFT-based TFRs so that the sensitivity of results to the TFR method could be checked.
