Flexibility and Negative Affect: Examining the Associations of Explanatory Flexibility and Coping Flexibility to Each Other and to Depression and Anxiety

Stone et al. (1991) and others highlight a number of limitations of extant self-report coping measures, particularly those based on the transactional model of coping. In a recent critique of self-report coping measures, Coyne and Racioppo (2000) argued that more situation-specific inventories would produce more meaningful results than traditional self-reported coping measures. The CSFI was designed to circumvent the limitations of coping measures that do not specify a time period for coping or the type of situations on which respondents are reporting and that have respondents rate their use of coping strategies in response to a single stressful situation. Finally, the CSFI was designed to provide a functional assessment of coping aims (i.e., the underlying purpose of coping) to circumvent difficulties associated with attempting to discern the function of coping based on the specific strategies or tactics that individuals endorse on self-report inventories.

The unselected college student sample had overall low levels of distress. On the BAI, 55% of the sample scored in the minimal range (0–7); 38% scored in the mild range (8–15); and 7% scored in the moderate range. On the BDI-II, as a total score, 84% scored in the minimal range (0–13); 12% scored in the mild range (14–19); and 4% scored in the moderate range. The restricted range of the manifest anxiety and depression symptom variables, and the subsequent latent negative affectivity variable, may have attenuated the relations between coping flexibility, explanatory flexibility, and negative affectivity in this sample.

Path analysis models with either BDI-II or BAI as dependent manifest variables, as well as a structural model in which the latent negative affect variable was assessed only with the BDI-II and BAI indicators, were also examined and produced comparable results. These results are available upon request of the corresponding author.

The Sobel test provides a test of the extent to which the mediator significantly carries the influence of an independent variable to a dependent variable (i.e., the indirect path). A significant Sobel z indicates that the mediator does significantly carry the influence from the IV to the DV. Freedman and Schatzkin (1992) provide a test of the difference between the adjusted and unadjusted regression coefficient (i.e., the change in the direct path with the inclusion of the mediator in the model). MacKinnon et al. (2002) provide evidence from a Monte Carlo study that compared 14 methods to testing mediation that the Sobel (1982) and Freedman and Schatzkin (1992) tests have the most power and the most accurate Type I error rates for statistically testing mediation in most situations.