Ridderprint

Fostering overseas success: A meta-analysis 107 Table 5.11: Estimated true correlations for direct effects. Success Adjustment Commitment Performance Retention Social support .24 .25 .33 .19 .21 Community domain .25 .30 .18 General friends .32 .22 Expatriates .13 Host country nationals .22 .28 .20 Family domain .32 .08 .19 .25 General family .25 .22 Spousal support .32 .08 .17 .25 Work domain .24 .35 .21 .19 Mentoring .27 .23 .19 .06 Peer support .27 .25 .17 .17 Supervisory support .24 .36 .26 .26 Cross-cultural training .16 .16 .06 .03 Logistical support .20 .22 .24 .20 POS .23 .52 .24 .26 POS adjustment .36 .28 .08 POS career .23 .32 .18 .28 POS financial .25 .27 .13 .19 Repatriation support .12 .18 Note: Numbers displayed are the estimated corrected true effect sizes (ρ) of the random effects models displayed in Tables 5.2 through 5.6. They vary in sample sizes, standard errors, and confidence intervals. Nonsignificant effects (95% CI includes zero) are in strikethrough whereas no effect size indicates a lack of independent samples ( k < 2). POS = perceived organizational support. 5.6 Discussion The primary purpose of this meta-analysis was to clarify and quantify the impact of social support on the success of international assignments of expatriates. Using data of 84 independent samples containing nearly 18,000 international assignments, we meta- analytically estimated the relationships between the social support provided from community, family, and work agents and four common criteria of expatriate success: cross-cultural adjustment, organizational commitment, performance, and retention. We validated the overall positive relationship between social support and these success criteria (Hypothesis 1). In spite of our expectations based on social exchange theories, we did not find stronger work-domain-specific reciprocation effects (Hypothesis 2). Rather, our results revealed strong crossover and spillover effects from family support on performance and retention, and from community support on performance and commitment. We also did not find moderating effects of mobility status – where we had expected that the effects of support by host country national agents and expatriate agents would differ (Hypothesis 3) – or of physical proximity – where we had expected that the support by host country agents would be stronger related to success (Hypothesis 4). Nevertheless, support by agents in the host country or agents with host country nationalities demonstrated considerable stronger associations with adjustment outcomes in our sampled studies. We confirmed that the effects of social support are stronger on proximal success criteria (adjustment and commitment) than on distal criteria (performance and retention; Hypothesis 5). Finally, we found that the relationship between social support and success was relatively robust to rater effects and publication bias.