156 Chapter 6 We proceeded by testing hypothesis 3a and 3b again while controlling for earlier levels of exhaustion (i.e., T1 exhaustion), functional limitations (i.e., T2 functional limitations), and absenteeism (i.e., T2 absenteeism). The conditional indirect effect for hypothesis 3a – stating that higher levels of T1 self-insight would strengthen the indirect relationship between T1 PVM and T3 functional limitations through T2 exhaustion – was no longer significant (estimate = -.05 [CI: -.20 to .01], p = .214). However, the direction and significance of the findings for hypothesis 3b were unaffected. The negative relationship between T1 PVM and T3 absenteeism through T2 exhaustion was again stronger at self-insight levels +1 SD above the mean (estimate = -.57, p < .05) than at self-insight levels -1 SD below the mean (estimate = -.08, p = .784). These findings provide additional support for hypothesis 3b. Finally, we tested moderated mediation hypothesis 4, which stated that higher levels of T1 self-insight strengthen the indirect relationship between T1 PVM and T3 creativity through T2 work engagement. We first tested this hypothesis without including the paths for earlier levels of the mediators (i.e., T1) and outcome variables (i.e., T2). The interaction effect between PVM and self-insight on work engagement was significant (estimate = .35, p < .01; see Figure 3. Moreover, in line with our hypothesis, the positive relationship between T1 PVM and T3 creativity through T2 work engagement was stronger at self-insight levels +1 SD above the mean (estimate = .20, p < .05) than at selfinsight levels -1 SD below the mean (estimate = .02, p = .652). After controlling for earlier levels of work engagement (i.e., T1 work engagement) and creativity (i.e., T2 creativity), the direction and significance of the findings for hypothesis 4 were unaffected. The positive relationship between T1 PVM and T3 creativity through T2 work engagement was again stronger at self-insight levels + 1SD above the mean (estimate = .07, p < .05) than at self-insight levels -1 SD below the mean (estimate = -.05, p = .158). These additional findings provide further support for hypothesis 4.
RkJQdWJsaXNoZXIy MTk4NDMw