Psychotherapy practice and research has been informed by the use of measures such as the Outcomes Questionnaire (OQ) that track changes in client symptoms over therapy. Clients seeking therapy primarily for anger problems do not score in the clinical range on the OQ, nor do they show improvement. To rectify this problem, we developed an Anger Management Outcomes Questionnaire (AMOQ). To devised this scale we rewordied the 18 items of the Anger Disorder Scale – Short Form (ADS-SF: DiGiuseppe & Tafrate, 2011) and added 6 items that reflected aspects of anger and aggression not on the ADS-SF, resulting in 24 items. The instructions ask subjects to complete a Likert scale based on their experiences in the last week.
We administered the AMOQ to 550 non-clinical participants recruited through Facebook and the internet to assess internal consistency, concurrent and discriminative validity. Subjects also completed the ADS-SF, the Beck Depression Inventory, and the Beck Anxiety Inventory. We collected AMOQ data on 55 clients who sought therapy for anger problems, or whose therapist indicated that anger was the client’s primary problem. For the angry clients the correlation between their OQ scores and their AMOQ was .25, suggesting that the two scales do not measure the same construct.
Factor analysis of the AMOQ yielded four factor, which we labeled 1) Anger-In, 2) Physical Aggression and Revenge, 3) Verbal Aggression and Coercion, and 4) Anger Targets. The total score and the four factor scores each demonstrated excellent internal consistency on Cronbach’s Alpha. The total and factor scores had larger correlations with the ADS-SF and the Dysfunctional Anger scale. Significant but small correlations occurred between the AMOQ scores with anxiety and depression scales. Also, clients starting anger treatment scored significantly higher on the AMOQ than the non-clinical sample. These results demonstrate good internal consistency, convergent and discriminative validity.
We continued to administer the AMOQ weekly to the clinical sample for 25 sessions. We used a mixed model regression analysis (Hierarchical Regression) to analyze the AMOC scores of these clients across sessions and for samples who received individual or group therapy. The results indicated that the AMOQ significantly decreased across sessions, and this effect was greater for clients receiving individual therapy comparted to group therapy.
We will add subjects to the clinical sample until the convention to increase power on this mixed model statistics, and to determine if the slop of recovery differs by higher scores at intake. We will add non-clinical subjects up to 800 to establish a clinically significant change cut scores and norms.
Raymond DiGiuseppe– Professor, St. John's University, Queens, New York
Danika Charles– St. John's University, Queens, New York
Rosina Pzena– Graduate Student, St. John's University, Queens
Briana Cheney– St. John's University, Queens, New York
William Taboas– Fordham University, Bronx, New York