What Happens If We Compare Chopsticks With Forks? The Impact of Making Inappropriate Comparisons in Cross-Cultural Research

• Published in: Journal of personality and social psychology 2008-11, Vol.95 (5), p.1005-1018
• Main Author: Chen, Fang Fang
• Format: Article
• Language: English
• Subjects: Bias; Biological and medical sciences; China; Cross Cultural Differences; Cross Cultural Psychology; Cross cultural studies; Cross-cultural analysis; Cross-Cultural Comparison; Crosscultural Analysis; Culture; Differences; Error of Measurement; Fundamental and applied biological sciences. Psychology; Human; Humans; Life Satisfaction; Mean; Measurement; Measurement Invariance; Methodology. Experimentation; Peoples Republic of China; Personality Inventory - statistics & numerical data; Psychology. Psychoanalysis. Psychiatry; Psychology. Psychophysiology; Psychometrics; Psychometrics - statistics & numerical data; Psychometrics. Statistics. Methodology; Quality of Life - psychology; Quantitative analysis; Reference Values; Regression analysis; Self Concept; Self Esteem; Social psychology; Statistical Regression; Test Bias; Test Validity; United States; United States of America
• ISSN: 0022-3514; 1939-1315
• DOI: 10.1037/a0013193

It is a common practice to export instruments developed in one culture to another. Little is known about the consequences of making inappropriate comparisons in cross-cultural research. Several studies were conducted to fill in this gap. Study 1 examined the impact of lacking factor loading invariance on regression slope comparisons. When factor loadings of a predictor are higher in the reference group (e.g., United States), for which the scale was developed, than in the focal group (e.g., China), into which the scale was imported, the predictive relationship (e.g., self-esteem predicting life satisfaction) is artificially stronger in the reference group but weaker in the focal group, creating a bogus interaction effect of predictor by group (e.g., self-esteem by culture); the opposite pattern is found when the reference group has higher loadings in an outcome variable. Studies 2 and 3 examined the impact of lacking loading and intercept (i.e., point of origin) invariance on factor means, respectively. When the reference group has higher loadings or intercepts, the mean is overestimated in that group but underestimated in the focal group, resulting in a pseudo group difference.