| As the number of expatriate professionals has increased exponentially, it is imperative that organizations give increased attention to the effectiveness of their expatriate employees. This is especially true within the development sector, where expatriates play a key role in the success of international non-governmental organizations (NGOs) working to provide healthcare and humanitarian aid within developing countries. The current study was designed to examine a proposed model of the environmental (perceived organizational support [host & parent]), affective commitment (host & parent), and individual (sociocultural adaptation & psychological well-being) variables that contribute to the effectiveness of expatriate professionals working for an international NGO within a foreign country. Expatriates were recruited using snowball techniques through a large research university in the Pacific Northwest as well as several international NGOs; recruitment was also conducted through social networking sites (e.g., Facebook, LinkedIn, expatriate listservs). Just over half of participants (N = 159) were female (59.1%); 62.9% were U.S. nationals; 37.1% were international to the U.S., representing 25 nationalities; 88.7% intended to complete their assignment. Latent variable modeling using SEM procedures was used to assess model fit. The fit of the a priori model was unsatisfactory (χ2[338, N = 159] = 657.59, p > .001; NFI = .827, CFI = .907, and RMSEA = .077) and a model trimming approach was used to derive a better fitting model. The final model displayed a good fit (χ2 [163; N = 159] = 280.258, p < .01, NFI = 0.884, CFI = 0.947, and RMSEA = 0.067). The final model provided a parsimonious and integrated model of work performance that is consistent with previously supported models of work performance and includes the following variables: (a) perceived organizational support from the host and parent organization, (b) affective commitment to the host organization, and (c) sociocultural adaptation to the host culture. Although several hypotheses were supported, there were several limitations: (a) sample size less than (N = 190) needed to ensure reproducible results, (b) lack of information on participants' organizational affiliations, (c) exploratory nature of the model generating approach, and (d) self-report nature of participants' work performance scores. |
