Abstract
I developed a Chinese version of the Aggression Questionnaire (CitationBuss & Perry, 1992) by translating scale items into Chinese and subjecting them to standard validation procedures. I used confirmatory factor analysis via structural equation modeling to compare several measurement models. Models based on CitationBuss and Perry's (1992) original four-factor (29-item) scale failed to replicate in the Chinese sample; however, the construct validity of CitationBryant and Smith's (2001) abridged version of the Aggression Questionnaire received strong overall support. The new 12-item scale demonstrated good fit to the data and adequate internal reliability. Evidence for criterion validity was provided by the scale's sensitivity to differing levels of aggression in males and females. Convergent and discriminant validity received partial support from the pattern of correlations with a measure of anger rumination. Linguistic and metric equivalence were supported by high correlation coefficients between scores on Chinese and English versions of the scale completed by bilingual Chinese on separate occasions. Consistent replications of these preliminary results across three independent samples suggest that the Chinese version of the Aggression Questionnaire may be useful for clinical assessment and cross-cultural research.
ACKNOWLEDGMENTS
This research was jointly funded by University of Hong Kong Seed Funding for Basic Research and a Competitive Earmarked Research Grant (HKU 7447/05H) awarded by Hong Kong Government's Research Grants Council. Copies of the Chinese Aggression Questionnaire can be obtained from J. P. Maxwell. I thank Cindy Sit, Chris Chow, and Eva Chu for assistance with translation; Edvard and Marie Rösiö for insightful discussions; and two anonymous reviewers for their helpful comments on earlier versions of the article.