Do Orientations to Happiness Mediate the Associations Between Personality Traits and Subjective Well-Being?

An alternative model was also analyzed for each outcome variable (see Hayes 2013, for an extended discussion of this process). More specifically, we examined alternative models where personality traits mediated the associations between orientations to happiness and subjective well-being in order to compare the fit of these models with our proposed models. Our results suggested that these alternative models did not fit the data as well as our proposed mediational models (i.e., orientations to happiness mediating the associations between personality traits and subjective well-being) because the confidence intervals for the alternative models were substantially closer to zero or actually included zero. For example, our strongest finding of the association between extraversion and life satisfaction through meaning in life yielded a confidence interval of .12–.41, whereas the strongest alternative model yielded a confidence interval of .05–.25. Furthermore, the magnitude of the effects in the alternative models were, at best, roughly half as strong as the proposed models. For example, the indirect effect of an orientation to meaning on life satisfaction through extraversion was ab = .14 as opposed to our proposed model of ab = .25 when extraversion was the predictor variable and meaning was the mediator. However, future research could benefit from testing other potential alternatives (e.g. life satisfaction mediating the association between personality and orientations to happiness).

Additionally, the three proposed models were re-examined excluding variables that did not uniquely predict mediator or outcome variables (e.g. Honesty-Humility). These analyses displayed results consistent with the model reported; however, they explained between .03 and .04 less variability in the outcome variables than the models provided. Furthermore, these simpler models over-exaggerate the extent to which any single personality variable accounts for variability in subjective well-being because they do not account for shared variability among the other personality variables.