Jokinen’s (2005) definition of global leadership competencies stated earlier guided our selection of study participants. We screened for participants who were leading across cultures, and working in many different industries. We were not concerned about different levels of responsibility. A wide variety of organizations were included, from mid-size to Fortune 100 organizations. There were 134 industries represented. Some 66% of respondents were male, 34% female (mean age = 43). Only global leaders were included in the study. Some 80 NASA International Project Managers and their international colleagues were included. We defined a global leader as one who is engaged in managing people across cultures either on international assignment or working …show more content…
Across the entire sample (N = 1,953), the six factor structure did not fit the data well, χ2 (1024) = 9398.87, CFI = .70, TLI = .68, RMSEA = .07, according to CFI and TLI (>.90) and RMSEA (<.08) fit standards suggested by MacDonald and Ho (2002). Based on these criteria, the six factor model did not fit the data well. In exploring the factor loadings of the six factor model it became clear items in the Adapting Socially, Even Disposition, and Demonstrating Creativity factors were not functioning as expected. In addition, one item from the Instilling Trust factor fit poorly with the other items under that factor, and was dropped from subsequent analyses. This left 23 items across three factors: Respecting Beliefs, Navigating Ambiguity, and Instilling Trust. The three-factor model, which included a higher-order “global leadership” factor, fit the data substantially better, χ2 (227) = 1608.48, CFI = .89, TLI = .88, RMSEA = .06, and approached acceptable fit according to MacDonald and Ho’s (2002) standards. In addition, each of the three factors loaded highly onto the higher order factor - Respecting Beliefs = 0.79, Navigating Ambiguity = 0.84, and Instilling Trust = 0.87—suggesting that each was a relatively strong indicator of the higher order construct of global leadership. We were unable to find support for Hypothesis 1 in that the proposed 10 factor model did not demonstrate acceptable model fit. However, we were able to identify a more parsimonious alternative model that demonstrated improved