Exploratory factor analysis and convergent validity of the Dundee Provocation Inventory

Background The prevalence and consequences of anger and aggression in people with intellectual disability (ID) are of great concern. It is essential that appropriate assessment tools are developed to aid formulation of treatments and to evaluate progress and outcomes. Method This study evaluates the Dundee Provocation Inventory (DPI), a 20-item assessment measure for anger provocation. A group of 114 participants were administered the DPI, and 62 of these were also administered the Novaco Anger Scale (NAS) and NAS Provocation Inventory (NAS-PI), two well-validated measures. Results Preliminary analysis revealed that the DPI correlated significantly with the NAS (r = .57) and NAS-PI (r = .77). The DPI had high internal consistency (alpha = .91) and moderate to high inter-item and item-to-total score correlations. Factor analysis revealed a 5-factor solution which accounted for 63% of the variance and was most easily interpreted. Conclusion The analysis suggests that the DPI is a suitable tool for assessing anger in people with ID. Further replication of the factor structure would be valuable.