Explaining variations in the findings of presenteeism research: A meta-analytic investigation into the moderating effects of construct operationalizations and chronic health.

The purpose of the present study is to theorize and test the moderating effects of two variables-the way presenteeism is operationalized and the presence of a preexisting chronic health condition-on the relationships between presenteeism and its antecedents (i.e., physical health, mental health, work factors, social factors, and personal factors). A meta-analysis of 116 studies (N = 301,402) investigated the impact of both moderator variables while controlling for the country of the sample and publication source. As expected, the magnitude of the relationships between presenteeism and its antecedents varied depending on the type of operationalization of presenteeism. Specifically, the average mean correlations reported in previous studies were larger when presenteeism was operationalized using both a behavior and an outcome (e.g., productivity loss stemming from attending work while ill) as compared with when presenteeism was operationalized as a behavior only (e.g., attending work while ill). Furthermore, we found that the associations between presenteeism and its antecedents were stronger for those workers with a preexisting chronic health condition (e.g., osteoarthritis), as compared with healthier workers. These findings have important implications for research and theory. In particular, they suggest that the way presenteeism is operationalized can artificially inflate the observed effect sizes between presenteeism and its antecedents. This is a significant contribution, as it may shape future measures of presenteeism. Theoretically, the findings are also important, as they provide a framework for understanding why some workers are more prone to presenteeism than others (e.g., because individuals with chronic health problems may be more resilient). (PsycINFO Database Record (c) 2018 APA, all rights reserved).