The three personality traits seemed to relate differentially to anxiety. In the direct effects model (including the traits and anxiety), only hardiness had a direct, significant path to anxiety. Although this model was not the best fit, these results are surprising, particularly given the correlations among hardiness, PGI, CSE, and anxiety. It may be that hardiness accounts for part of the significant correlations of PGI and CSE with anxiety; if they were by themselves in a direct effects model, significant paths from PGI and CSE to anxiety might exist.