Bias and equivalence provide a framework for methodological aspects of cross-cultural studies. Bias is a generic term for any systematic errors in the measurement that endanger the comparability of cross-cultural data; bias results in invalid comparative conclusions. The demonstration of equivalence (i.e., absence of bias) is a prerequisite for any cross-cultural comparison. Based on the source of incomparability, three types of bias, namely construct, method, and item bias, can be distinguished. Correspondingly, three levels of equivalence, namely, construct, metric, and scalar equivalence, can be distinguished. One of the goals in cross-cultural research is to minimize bias and enhance comparability. The definitions and manifestations of these types of bias and equivalence are described and remedies to minimize bias and enhance equivalence at the design, implementation, and statistical analysis phases of a cross-cultural study are provided. These strategies involve different research features (e.g., decentering and convergence), extensive pilot and pretesting, and various statistical procedures to demonstration of different levels of equivalence and detections of bias (e.g., factor analysis based approaches and differential item functioning analysis). The implications of bias and equivalence also extend to instrument adaptation and combining etic and emic approaches to maximize the ecological validity. Instrument choices in cross-cultural research and the categorization of adaptations stemming from considerations of the concept, culture, language, and measurement are outlined. Examples from cross-cultural research of personality are highlighted to illustrate the importance of combining etic and emic approaches. The professionalization and broadening of the field is expected to increase the validity of conclusions regarding cross-cultural similarities and differences.
A Giant of Cultural Research: Seeing Further from the Shoulders of Kwok Leung
