Erratum to: Cogn Ther Res DOI: 10.1007/s10608-012-9506-z

A veteran with sub-threshold PTSD was erroneously included in our sample of veterans with a current PTSD diagnosis. Upon correcting the error and rerunning analyses, we found that while the magnitudes of relationships among most variables were not meaningfully altered, some statistical results were slightly changed (see Tables 1, 2). In the new analyses, PTSD numbing no longer partially mediates the relationships between all PTSD clusters and trait anger. Instead, numbing partially mediates the relationships between the reexperiencing cluster and trait anger, and between the avoidance cluster and trait anger. Hyperarousal now has a significant indirect effect on trait anger through numbing symptoms. Additionally, PTSD dysphoria symptoms now partially mediate the relationships between all PTSD clusters and trait anger.

Table 1 Standardized path estimates, total effects, and indirect effects for analyses with current MDD as the mediator
Table 2 Standardized path estimates, total effects, and indirect effects for analyses with numbing and dysphoria clusters as mediators

The corrected results continue to support the main findings of our study; however, contrary to our original discussion, we can now conclude that numbing and dysphoria have a similar impact on PTSD cluster–anger relationships.