Questions on 9/11 included the type of exposure to the attacks. From these, we used an ordinal NESARC 9/11 variable with four exposure levels, as described elsewhere (Henriksen et al., 2010): (1) No exposure; (2) Indirect exposure through media, news, etc.; (3) Had a close friend or family member injured or killed; and (4) Direct exposure (respondent at the scene of the terrorist attack and/or injured). If a respondent experienced 9/11 in more than one way, they were coded at the highest level of exposure they experienced. The four levels were dummy-coded for comparisons of different exposure levels.
