Some limitations of this study deserve mention. First, the sample of depressed patients in the current study consisted of patients who mainly had a unipolar, non-melancholic depression with little comorbidity. It is therefore unclear to what extent the findings generalize to other diagnostic subtypes of depression or to patients with comorbid anxiety disorders. Second, the novelty P3 component overlaps the P3b component, which leaves open the possible contribution of P3b to group differences in the mean integrated amplitude of the novelty P3. The use of multivariate techniques, such as principal components analysis (PCA), could aide in identifying and measuring these separate P3 subcomponents. An independent replication and extension of this study using a larger electrode array (67 channels) applied PCA to reference-free Laplacian transformations of ERPs during the novelty oddball task (Tenke et al., 2009). This analysis identified an early mid-central source (245-ms peak latency) that was unique to novel stimuli and was markedly reduced in depressed patients when compared to healthy controls. Again, group differences were less evident for later sources corresponding to P3b. This supports the hypothesis that
