The generalizability of the Buss-Perry Aggression Questionnaire

Aggressive and hostile behaviours and anger constitute an important problem across cultures. The Buss-Perry Aggression Questionnaire (AQ), a self-rating scale was published in 1992, and has quickly become the gold-standard for the measurement of aggression. The AQ scale has been validated extensively, but the validation focused on various narrowly selected populations, typically, on samples of college students. Individuals, however, who are at risk of displaying aggressive and hostile behaviours; may come from a more general population. Therefore, it is important to investigate the scale's properties in such a population. The objective of this study was to examine the factorial structure and the psychometric properties of the AQ scale in a nationally representative sample of the Hungarian adult population. A representative sample of 1200 subjects was selected by a two-step procedure, The dimensionality and factorial composition of the AQ scale was investigated by exploratory and confirmatory factor analyses. Since spurious associations and increased factorial complexity can occur when the analysis fails to consider the inherently categorical nature of the item level data, this study, in contrast to most previous studies, estimated the correlation matrices subjected to factor analysis using the polychoric correlations. The resulting factors were validated via sociodemographic characteristics and psychopathological scales obtained from the respondents. The results showed that based on the distribution of factor loadings and factor correlations, in the entire nationally representative sample of 1200 adult subjects, from the original factor structure three of the four factors (Physical and Verbal Aggression and Hostility) showed a good replication whereas the fourth factor (Anger) replicated moderately well. Replication further improved when the sample was restricted in age, i.e. the analysis focused on a sample representing the younger age group, comparable to that used in the original Buss-Perry study. Similar to the Buss-Perry study, and other investigations of the AQ scale, younger age and male gender were robustly related to physical aggression. In addition, level of verbal aggression was different between the two genders (with higher severity in mates) whereas hostility and anger were essentially the same in both genders. In conclusion, the current study based on a representantive sample of adult population lends support to the use of the AQ scale in the general population. The authors suggest to exclude from the AQ the two inverse items because of the low reliability of these items with regard to their hypothesized constructs. Copyright (c) 2007 John Wiley & Sons, Ltd.