Interest in presenteeism, attending work while ill, has flourished in light of its consequences for individual well-being and organizational productivity. Our goal was to identify its most significant causes and correlates by quantitatively summarizing the extant research. Additionally, we built an empirical model of some key correlates and compared the etiology of presenteeism versus absenteeism. We used meta-analysis (in total, K=109 samples, N=175,965) to investigate the correlates of presenteeism and meta-analytic structural equation modeling to test the empirical model. Salient correlates of working while ill included general ill health, constraints on absenteeism (e.g., strict absence policies, job insecurity), elevated job demands and felt stress, lack of job and personal resources (e.g., low support and low optimism), negative relational experiences (e.g., perceived discrimination), and positive attitudes (satisfaction, engagement, commitment). Moreover, our dual process model clarified how job demands and job and personal resources elicit presenteeism via both health impairment and motivational paths, and they explained more variation in presenteeism than absenteeism. The study sheds light on the controversial act of presenteeism, uncovering both positive and negative underlying mechanisms.The greater variance explained in presenteeism as opposed to absenteeism underlines the opportunities for researchers to meaningfully investigate the behavior and for organizations to manage it.Keywords: presenteeism; absenteeism; meta-analysis; health; demands. The term presenteeism has been used in several ways in the literature over the years, and Johns (2010) reviewed nine distinct definitions of the concept (see also Wężyk & Merecz, 2013).However, recent scholarly treatment has converged on two main definitions. The first of these is attending work while ill (Aronsson, Gustafsson, & Dallner, 2000). The second definition, which dominates the occupational medicine literature, is productivity loss stemming from attending while ill (Turpin et al., 2004). Although affirming the importance of the latter phenomenon, Johns (2010Johns ( , 2012 questioned this definition for conflating cause with effect and pre-empting the study of the causes of going to work ill. In the meta-analyses to be reported here we restrict the purview to the causes and correlates of the act of presenteeism-going to work ill. Most research on productivity loss has concentrated on its association with various medical conditions. Perhaps the most interesting thing about presenteeism is that it represents a muchoccupied but only recently studied state between being absent (and ostensibly exhibiting no productivity) and fully productive work engagement. Presentees will vary in their productivity due to a host of personal and contextual circumstances, such as the exact nature of their health Presenteeism correlates: A meta-analysis 4 problems. Nonetheless, they are unlikely to be fully engaged and fully productive. Hence, the causes and corre...

A debate has arisen out of the need to understand true intervention outcomes in the social sciences. Traditionally, the randomized, controlled trial (RCT) that answers the question of 'what works' has been considered the gold standard. Although RCTs have been favoured in organizational intervention research, there has been an increasing interest in understanding the influence of context and intervention processes on the outcomes of such interventions. In the present critical essay, we question the suitability of RCTs and meta-analyses to evaluate the effectiveness of organizational interventions and we suggest that realist evaluation that seeks to answer the questions of what works for whom in which circumstances may present a more suitable framework. We argue that examining the content and process mechanisms through which organizational interventions are effective, and the conditions under which these are triggered, will enable us to better understand how interventions achieve the desired outcomes of improved employee health and well-being. We suggest that organizational intervention content and process mechanisms may help bring about the desired outcomes of improved employee health and well-being and that contextual factors determine whether these mechanisms are triggered. KeywordsCMO-configurations, context-mechanism-outcome configurations, critical essay, metaanalysis, organizational interventions, randomized controlled trial, realist evaluation, realist synthesis What works for whom?What works for whom?The randomized, controlled trial has been perceived as the gold standard for evaluating interventions (Guyatt et al., 1995). The RCT approach employs a successionist approach to causation, i.e. that randomization holds the context constant, there are no differences at baseline between the intervention and the control group, and outcomes can be inferred from comparing those exposed to the intervention to those not exposed to the intervention. When intended outcomes can be observed statistically above and beyond outcomes in the control group, it is assumed that these outcomes can be attributed to the intervention. Implicit in this line of thinking is that intervention outcomes can be aggregated What works for whom?In the field of organizational interventions the RCT design has been met with criticism (Nielsen, 2013a). Organizational interventions can be defined as planned, behavioural, theory-based actions that aim to improve employee health and well-being through changing the way work is designed, organized and managed (e.g. Nielsen, 2013a;Richardson and Rothstein, 2008). In the present essay, we argue that realist evaluation (Pawson, 2013;Pawson and Tilley, 1997) may offer an opportunity to develop an integrated context, process and outcome evaluation framework that may advance our theoretical understanding of which elements of organizational interventions may be effective and in which conditions we can expect positive outcomes. Realist evaluation offers a way to conduct rigorous, theory-based analyse...

Background: This study investigated the association between presenteeism and the perceived availability of social support among hospital doctors in China. Methods: A questionnaire was administered by doctors randomly selected from 13 hospital in Hangzhou China using strati ed sampling. Logit model was used for data analysis. Results: The overall response rate was 88.16%. Among hospital doctors, for each unit increase of the perceived availability of social support, the prevalence of presenteeism was decreased by 8.3% (OR=0.91, P=0.000). In particular, if the doctors perceived availability of appraisal support, belonging support and tangible support as su cient, the act of presenteeism was reduced by 20.2% (OR=0.806, P=0.000) 20.4% (OR=0.803, P=0.000) and 21.0% (OR=0.799, P=0.000) respectively with statistical differences. Conclusion: In China, appraisal support, belonging support and tangible support, compared to other social support, had a stronger negative correlation with presenteeism among hospital doctors. The bene ts of social support in alleviating doctors' presenteeism warrant further investigation.

Among job attitudes, overall job satisfaction has received the greatest attention in organizational research and it has frequently been suggested as the key factor influencing employees' performance. Although it reflects individual experiences, job Many people spend a significant amount of time in their workplace and the feelings of work-related satisfaction or dissatisfaction contribute to overall quality of life and psychological well-being (Judge and Watanabe 1993;Wright et al., 1999).Beyond the value of positive feelings for the individual, the benefits for organizations have been widely investigated, stressing the impact of job satisfaction on several organizational outcomes (e.g., Judge and Kammeyer-Mueller, 2012;Spagnoli et al., 2012). Moreover, the link between job satisfaction and job performance has long been of interest to organizational psychologists and several studies have suggested that job satisfaction is a key factor influencing productivity and job performance (Judge et al., 2001; Riketta, 2008).Up to now, job satisfaction has been studied mainly at the individual level, focusing on employees' characteristics like self-efficacy, core self-evaluations, and dispositional affect (Fernandez-Ballesteros et al., 2002; Judge and Kammeyer-Mueller 2012). A few studies have related work resilience and job satisfaction (Larson and Luthans, 2006;Youssef and Luthans, 2007), showing that individuals with higher levels of resilience are more likely to positively adapt and successfully bounce back from negative events in the workplace, and this can enhance their job satisfaction. However, these few studies are mostly correlational and cross-sectional, making difficult to establish causal relationships. Although job satisfaction reflects an evaluation of individual experiences, it is also likely to be affected by the attributes of the context in which the individual operates (Ostroff, 1992(Ostroff, , 1993. Social environment variables, such as relationships with coworkers and supervisors, are closely related to job satisfaction and predict satisfaction levels above and beyond characteristics of the work itself (Judge RUNNING HEAD: Predicting job satisfaction and performance 3 and Kammeyer-Mueller, 2012; Morgeson and Humphrey, 2006). In this regard, Borgogni and colleagues (Alessandri et al., 2014; Borgogni et al., 2011a; Borgogni et al., 2010b) introduced the concept of "Perceptions of Social Context" (PoSC 1 ), defined as the individual's perceptions of the more relevant social constituents internal to the organization (i.e., top management, immediate supervisor, and colleagues). At the aggregated level, PoSC could work as a broad concept reflecting the overall work-unit perception of the social environment.In light of the paucity of studies investigating the interplay of individual and group variables in shaping job satisfaction, it seems imperative to explore its antecedents from a multilevel perspective. To describe the interrelationships among variables measured at different levels (i.e.,...

Because of a variety of access limitations, self-reported absenteeism from work is often employed in research concerning health, organizational behavior, and economics, and it is ubiquitous in large scale population surveys in these domains. Several well established cognitive and social-motivational biases suggest that self-reports of absence will exhibit convergent validity with records-based measures but that people will tend to underreport the behavior. We used meta-analysis to summarize the reliability, validity, and accuracy of absence self-reports. The results suggested that self-reports of absenteeism offer adequate test–retest reliability and that they exhibit reasonably good rank order convergence with organizational records. However, people have a decided tendency to underreport their absenteeism, although such underreporting has decreased over time. Also, self-reports were more accurate when sickness absence rather than absence for any reason was probed. It is concluded that self-reported absenteeism might serve as a valid measure in some correlational research designs. However, when accurate knowledge of absolute absenteeism levels is essential, the tendency to underreport could result in flawed policy decisions.