eling (SEM) using the partial least squares (PLS) method was used to answer the research questions. In an external model, PLS analysis is used to estimate latent variables (LVs) based on the shared variance of observed variables, using the principal-component weights of the observed variables. The shift in each indicator indicates the extent of its influence on a given LV, resulting in the best possible combination of weights for predicting the LV while accounting for observed variables. We first assumed that all of the hypothesized relations were linear, and used the software package SmartPLS to test the model using standard linear PLS analysis. The results of the preliminary analysis failed to support some of the hypotheses. However, an examination of bivariate data plots suggested the presence of asymmetric effects. Further correlation analysis of split samples also revealed nonlinear relationships. For instance, charter value (CV) may have two distinct effects: low and high levels of CV may lead to lower performance than medium levels of CV, while low to medium CV can motivate individuals to achieve their highest performance (Figure 6-1). Accordingly, using WarpPLS? [151] and the guidelines developed by Kock and Mayfield [152], the quality of the measures was assessed by inspecting item-to-total correlations.
reference(omitted)
