부산대학교 도서관

Background: Reports of a meaningful relationship between mental health-related conditions and work productivity measures are relatively common. These, however, are frequently examined for their linearity while ignoring untapped, and potentially rich, non-linear associations. Methods: Following a serendipitous finding of a curvilinear relationship between workplace presenteeism (lowered productivity while at work) and depression, an investigation was undertaken of the association between worklife prevalence measures of presenteeism (measured by the W.H.O. Health & Work Performance Questionnaire) and lifetime prevalence of twelve psychosocial vulnerabilities, encompassing mental health, mental health-related, and addictive conditions. Linear and quadratic (U-shaped) functions were calculated across the "relative" presenteeism measure (self vs. other workers) for each of the 12 conditions. Results: A visual analysis revealed a U-shaped graphic function in all conditions, and excepting anxiety all were statistically significant. In general, increases beyond the lowest ("poorest") level of self-reported comparative productivity were associated with increases in psychosocial stability, but only as far as deemed equality. Beyond that, increases in self-confidence resulted in a reversal, thus returning to a higher level of vulnerability for the condition in question. A cursory scan of five relevant journals indicated that non-linear analyses were often possible, but rarely carried out. Conclusions: This has informative value for our conceptualization of overconfidence, and it begs the question of whether an over-reliance on linear measures has caused us to overlook important curvilinear human relationships. The inclusion of analyses of non-linear functions is suggested as a matter of course for future studies. [ABSTRACT FROM AUTHOR]