The present study was not without limitations. First, while we reduced the number of significance tests for the styles analyses by focusing only on styles related to Neuroticism and Conscientiousness, we did not use a formal correction to control for Type I errors. We chose this approach because, in the presence of a large number of predictors, standard methods of correcting for Type I errors (e.g., the Bonferroni) overcorrect and lead to a preponderance of Type II errors. While Type I errors are serious in clinical intervention studies, as they can lead to the implementation of costly programs that may be ineffective, Type II errors are more serious when they impair capacity to identify at-risk individuals. Given the significant unmet need for services among the depressed elderly, statistical procedures that further obscure risk detection are highly questionable. To alleviate concerns about Type I errors, however, we did note which findings were significant at a more stringent p-value (p < .01) and focus discussion on findings that were theoretically meaningful. Second, the present sample was mostly white and unique in its
