153 Proactive Vitality Management Among Employees with Chronic Liver Disease respective latent factors fit the data reasonably well (CFI = .90, IFI = .90, TLI = .88, RMSEA = .08, SRMR = .09). Moreover, model fit improved when taking into account the three-dimensional structure of work engagement – i.e., the subdimensions vigor (3 items), dedication (3 items), and absorption (3 items) (CFI = .93, IFI = .94, TLI = .92, RMSEA = .06, SRMR = .07). All items loaded significantly on their respective latent factors (all factor-loadings > .62; p’s < .05). Hypotheses Testing The Health Impairment Process According to hypothesis 1, T1 PVM is negatively related to T3 functional limitations (hypothesis 1a) and T3 absenteeism (hypothesis 1b), through T2 exhaustion. We started with analyzing the simple model without including the paths for earlier levels of the mediators (i.e., T1) and outcome variables (i.e., T2). The model involving the health impairment process fit reasonably well to the data (CFI = .91, IFI = .92, TLI = .90, RMSEA = .08, SRMR = .09). In line with our predictions, the results showed significant paths from T1 PVM to T2 exhaustion (estimate = -.39, p < .01), from T2 exhaustion to T3 functional limitations (estimate = .59, p < .01), and from T2 exhaustion to T3 absenteeism (estimate = .36, p < .01). See Table 2 for all direct path coefficients. The indirect relationship between T1 PVM and T3 functional limitations through T2 exhaustion was negative and significant (AMOS bootstrapping estimate = -.24 [CI: -.40 to -.12], p < .01), and the indirect relationship between T1 PVM and T3 absenteeism through T2 exhaustion was negative and significant as well (AMOS bootstrapping estimate = -.14 [CI: -.28 to -.05], p < .01). Overall, these results provide initial support for hypotheses 1a and 1b. To control for earlier levels of the mediator and outcome variables, we added T1 exhaustion, T2 functional limitations, and T2 absenteeism to our structural equation model. Doing so did not affect the direction and significance of the results. T1 PVM was still negatively and significantly related to T3 functional limitations through T2 exhaustion (AMOS bootstrapping estimate = -.11 [CI: -.26 to -.03], p < .05). Moreover, T1 PVM was still negatively and significantly related to T3 absenteeism through T2 exhaustion (AMOS bootstrapping estimate = -.09 [CI: -.21 to -.04], p < .01). These findings provide further support for hypotheses 1a and 1b. 6
RkJQdWJsaXNoZXIy MTk4NDMw