89 Proactive Vitality Management, Work Engagement, and Creativity Each regression analysis was built on the basis of four nested models introducing successively the intercept (Null model), the lagged control variable and the predictor (Model 1), the two moderators (Model 2), and the two hypothesized interaction effects (Model 3; see Tables 2 and 3). Prior to the analyses, we tested whether the slope between the independent variables (e.g., proactive vitality management and work engagement) and the dependent variables (e.g., work engagement and creativity, respectively) varied across respondents. The slope variance was significant in both cases, justifying the introduction of between-level variables (i.e., goal orientation scales) so as to test crosslevel interaction effects on the week-level dependent variables. RESULTS Descriptive Statistics Means, standard deviations, and intercorrelations for all study variables are presented in Table 1. To calculate intercorrelations, week-level variables were aggregated over the five weeks. TABLE 1 Means, standard deviations and inter-correlations for the study variables (N = 107 employees and N = 501 occasions). M SD 1 2 3 4 5 Trait-level 1. Learning goal orientation 4.51 .72 - 2. Performance avoid goal orientation 5.05 .92 .25** - Week-level 3. Proactive vitality management 5.24 .54 .27** .09 - .44** .28** 4. Work engagement 5.00 .85 .23* .21* .55** - .47** 5. Creativity 4.37 1.02 .26** .10 .49** .53** - Note. Correlations below the diagonal refer to the between-level of analysis, while correlations above the diagonal refer to the within-level of analysis; means and standard deviations refer to the between-level of analysis. *p < .05, **p < .01 4
RkJQdWJsaXNoZXIy MTk4NDMw