4. Discussion
The present study examined the psychometric properties of three
versions of the AQ, using ESEM and the more typical CFA. The results
suggest that both short-forms are viable but are inconclusive in terms
of which scale is optimal.
Using CFA, the AQ demonstrated unsatisfactory model fit, the AQ-SF
demonstrated borderline good fit (and superior fit in comparison to the
AQ and BAQ) for the four-factor and hierarchical models, and the BAQ
demonstrated acceptable fit for the four-factor model. In contrast,
using ESEM, all three scales demonstrated fair to good model fit for
the four-factor model, with the BAQ demonstrating best fit. Regardless
of whether CFA or ESEM was employed, the findings support the fourfactor
model of aggression in the AQ-SF and the BAQ.
In terms of loading onto their hypothesised factors, problems were
evident with the Hostility items, where four loaded substantively