Multilevel cross-cultural training outcomes and self-enhancement or strategic modesty through self-reported outcomes: a meta-analysis

Abstract

Existing meta-analyses of cross-cultural training (CCT) outcomes have been insightful but have not utilized a systematic multilevel assessment framework to evaluate CCT-outcomes. We conduct a meta-analysis of CCT outcomes using Kirkpatrick’s multilevel outcomes – reaction, learning, behavior, and results – for training evaluation. The analyses are based upon 319 relationships (k = 319) from 123 different studies (K = 123) with a combined sample of 40,711 (N = 40,711). Results using hierarchical robust variance estimate (RVE) models illustrate a small positive overall relationship between CCT and targeted reaction ( r ¯ =0.243, p < 0.05), learning ( r ¯ =0.294, p < 0.05), behavior ( r ¯ =0.244, p < 0.05), and results ( r ¯ =0.259, p < 0.05) outcomes. Additionally, learning-level outcomes may be difficult for external observers to accurately assess, so trainees may use self-reports of learning-level outcomes to engage in self-enhancement. Alternatively, results-level outcomes may be easier for observers to assess, so trainees may engage in strategic modesty while self-reporting results-level outcomes. Consistent with these ideas, using RVE models for meta-regression we find a positive interaction effect between learning-level outcomes and the use of self-reporting ( β = 0.201, p < 0.05) and a negative interaction effect between results-level outcomes and the use of self-reporting ( β = − 0.322, p < 0.01).