63 Proactive Vitality Management and Creative Work Performance TABLE 1 Descriptive Statistics and Correlations Variables M SD 1 2 3 4 5 6 7 Person-level 1. Self-insight 4.61 .64 - 2. Social support for creativity 5.27 1.07 .06 - 3. Proactive personality 3.55 .52 .16*** .13** - 4. Creative requirement 2.92 1.07 -.01 .14*** .27*** Week-level 5. Proactive vitality management 5.01 .81 .19** .18** .18*** .07 - 6. Creative work performance 3.08 .63 .05 .16*** .36*** .27*** .44*** - 7. Job autonomy 3.72 .73 .18** .23*** .22*** .10* .38*** .39*** - Note. N = 242 employees and n = 610 observations. Self-insight was scored on a 6-point scale, social support for creativity and proactive vitality management were measured using a 7-point scale and proactive personality, creative requirement, job autonomy and creative work performance were scored on a 5-point scale. *p < .05, **p < .01, ***p < .001 Multilevel Confirmatory Factor Analyses Prior to testing our hypotheses, we used Mplus software (Muthén &Muthén, 1998 - 2012) to conduct several relevant multilevel confirmatory factor analyses (MLCFAs). First, to examine the measurement model and check for construct validity and independence of our variables, we tested a measurement model containing four factors: Creative work performance (five items); Proactive vitality management (eight items); Self-insight (eight items); and social support for creativity (three items). The multilevel measurement model in which all items of all the variables in our model loaded on their respective latent factors fit the data well (CFI = .93, TLI = .91, RMSEA = .06, SRMR within = .05, SRMR between = .06). In addition, all factors had significant factor loadings (p < .01). Second, we wanted to test thoroughly whether we could empirically distinguish the predictor in our model (proactive vitality management) from the outcome (creative work performance). Therefore, we conducted two multilevel confirmatory factor analyses (MLCFAs) to compare a model in which the items of each construct load on their own respective latent factor to a model in which all items load on one overall latent factor. The model in which the indicators of the two constructs loaded on two separate factors had a good fit to the data (CFI = .93, TLI = .92, RMSEA = .06). Moreover, this model 3
RkJQdWJsaXNoZXIy MTk4NDMw