29 Conceptualization and Measurement of Proactive Vitality Management for it to be used conveniently. So, while we deliberately started out with a relatively large pool of items to empirically answer the question which items functioned best together in terms of their loadings, one of our goals was to significantly reduce the number of items. In the first analysis, SPSS extracted three factors based on their Eigenvalues (> 1). However, we noted that the first factor had an Eigenvalue (7.9) that was considerably higher than the other two factors (1.7 and 1.2, respectively). Only one item had a considerable loading on factor three, so we excluded this item/factor. In addition, the second factor did not make theoretical sense, i.e., it overlapped with the first factor regarding content. In the subsequent analysis, two items had high cross loadings on the second factor in the factor solution, so we excluded these items as well. In a further iterative process, two subsequent analyses were performed in which three more items were excluded, using the same criteria. The remaining twelve items loaded on one single factor. However, in order to achieve our goal and facilitate efficient use of the scale, we performed a content analysis and finally decided to exclude four more items that did not add unique, meaningful information to the scale. We were able to exclude these redundant items without compromising construct coverage and face validity (i.e., representation of all facets of the PVM construct). For example, one item was “I make sure that I can concentrate well on my work”, which is highly similar to “I make sure that I can focus well on my work”. In this case, we excluded the former item because it had a lower loading on the latent factor. The eight remaining items together formed one overall factor that is representative of the proactive vitality management construct. The factor had an Eigenvalue of 4.12 and explained 51.5% of the variance. The factor-loadings of the items ranged from .67 to .78, and Cronbach’s alpha of the eightitem scale was α = .86. The total general-level sample (N = 813) was used to calculate means and standard deviations of the items. The eight PVM items and their descriptive statistics can be found in Table 1. 2
RkJQdWJsaXNoZXIy MTk4NDMw