Key Studies

Smith, D. C., Furlong, M., Bates, M., & Laughlin, J. D. (1998). Development of the  multidimensional school anger inventory for males. Psychology in the Schools, 35, 1-15. doi:10.1002/(SICI)1520-6807(199801)35:1<1::AID-PITS1>3.0.CO;2-U

Drawing upon research which conceptualizes anger as a multidimensional construct including three associated components -- anger experience (affective anger), hostility (anger cognitions), and anger expression (aggression, assertion, and withdrawal) -- the preliminary development of a Multidimensional School Anger Inventory (MSAI) for adolescents is described. This scale is a modification and extension of the School Anger Inventory and was developed to assess the affective, cognitive, and expressive aspects of anger using items having school-relevant content. Data were collected through personal interviews of 202 males from three different schools: School 1 included general education students in a parochial school in grades 6 through 12; School 2 included students attending general education or mainstreamed special education classes at a public intermediate school; and School 3 included students participating in a public day treatment program for youths with serious emotional disturbance. Scale development is discussed focusing on item development and scale refinement through item and factor analyses. Four factors were identified that accounted for 43.3% of the common variance. Anger Experience, Cynical Attitudes, and Anger Expression were identified as major clusters with the anger expression items bifurcating into Destructive Expression and Positive Coping components. The resulting 31-item scale has strong psychometric qualities and appears to have promise for use in research, treatment planning, and outcome evaluations.

Furlong, M. J., & Smith, D. C. (1998). Raging Rick to Tranquil Tom: An empirically based  multidimensional anger typology for adolescent males. Psychology in the Schools, 35, 229-245.  Doi:10.1002/(SICI)1520-6807(199807)35:3<229::AID-PITS4>3.0.CO;2-I 

Anger-related problems among children and youth are of considerable interest to school psychologists and others concerned about the prevention and reduction of school violence. Both research and clinical efforts, however, have suffered from a lack of conceptual clarity with regard to anger, an emotion that is highly correlated with aggressive behavior. This article reviews efforts to categorize anger-related patterns among youth and proposes an empirically derived typology for males based upon a multidimensional conceptualization of anger in school settings. Utilizing a two-stage clustering procedure (hierarchical and K-means procedures), six anger preference styles were described for a sample of 200 participants in grades 6-12. As a heuristic to aid in the interpretation of the six styles, they were labeled Raging Rick (Cluster 1: Extreme Anger subtype), Bitter Bill (Cluster 2: Cynical subtype), Dynamic Don (Cluster 3: Impulsive subtype), Sociable Sam (Cluster 4: Prosocial subtype), Suppressive Sal (Cluster 5: Low Arousal-Low Coping subtype), and Tranquil Tom (Cluster 6: Low Arousal-High Coping subtype). External validity of the resultant groupings was evaluated by comparing subgroup profiles on both teacher and self-reported ratings of anger and aggression within school settings. This typology of anger-related styles has implications for diagnosis, prevention programs, and treatment planning within school settings.

Furlong, M. J., Smith, D. C., & Bates, M. (2002). Further development of the Multidimensional  School Anger Inventory: Construct validation, extension to female adolescents, and  preliminary norms. Journal of Psychoeducational Assessment, 20, 46-65. doi: 10.1177/073428290202000104

The responses of 1,166 students in Grades 9 through 12 to the Multidimensional School Anger Inventory (MSAI) were used in four analyses to further examine and extend its psychometric properties. In Analysis 1, the original four-factor structure (Anger Experience, Hostility, Destructive Expression, and Positive Coping) of the MSAI was validated with a subsample of males and extended to a sample of females. Analysis 2 used another nonoverlapping subsample to further improve the MSAI by exploring the utility of adding exploratory items to the two anger expression subscales (Destructive Expression and Positive Coping). Based upon this analysis, four items were added to Destructive Expression and one item was added to Positive Coping. A principal components analysis once again replicated the four-factor structure. Analysis 3 was completed using a third nonoverlapping subsample of males and females to cross-validate the four-factor structure of the refined MSAI. Given the robustness of the MSAI factor structure across nonoverlapping subsamples and its comparability for males and females, Analysis 4 combined data for all 1,166 students to develop preliminary norms for the MSAI. Test-retest data for 508 students found stability coefficients to be moderate (.56 to .62) across the four subscales. Uses of the MSAI for school-based assessment and program evaluation and avenues for further development are discussed.

Boman, P. (2003). Gender differences in school anger. International Education Journal, 4, 71-77. Available from,

This study examined gender differences in the affective, behavioural, and cognitive components of anger in 102 students completing their first year of high school. Results supported not only the hypothesis that girls and boys do not differ in their experience (affective) of anger but also the belief that girls are more likely to express positively (behavioural) their anger than boys. Additionally, results supported the expectation that boys are more hostile (cognitive) towards school than girls. Suggestions for future research and the relevance of the findings for schools are also addressed

Boman, P., Curtis, D., Furlong, M. J., & Smith, D. C. (2006). Cross-validation and Rasch analyses  of the Australian version of the Multidimensional School Anger Inventory-Revised. Journal of Psychoeducational Assessment, 24, 225-242.

The construct validity of the Australian version of the Multidimensional School Anger Inventory-Revised (MSAI-R) was examined using exploratory factor analysis (EFA), Rasch analysis, and confirmatory factor analysis (CFA) on a sample of 1,400 Australian students enrolled in Years 8 through 12. The EFA revealed a strong replication of the MSAI-R's internal consistency and structure with the original four MSAI-R factors identified: Anger Experience, Hostility, Destructive Expression, and Positive Coping. However, the Rasch analysis revealed that some additional items could be added to the destructive anger expression and positive coping subscales to enhance their reliability. A CFA was also conducted to compare the four-factor and a one-factor model. The results indicated a good model fit for the four-factor solution apart from one index. It is concluded that the results lend further support to the MSAI-R's construct validity and support its use as a general survey and evaluation research instrument.

Boman, P., Smith, D. C., & Curtis, D. (2003). Effects of pessimism and explanatory style on  development of anger in children. School Psychology International, 24, 80-94.  Available from,

 In this project 102 students in their first year of high school responded to questionnaires assessing their levels of dispositional optimism and pessimism, explanatory style and anger in relation to the school setting. A Partial Least Squares (PLS) path analysis examined relationships among these variables. Male students with helpless explanatory styles were more likely to resort to destructive school behaviour. Male students with a pessimistic disposition were also more likely to report higher levels of school hostility and destructive school behaviour. For females, helpless explanatory style and dispositional pessimism were related but the overall level of anger intensity did not appear to relate to destructive and aggressive behaviour. Overall the results suggest that anger manangement programs focusing on cognitive re-structuring and related strategies can be a powerful means for reducing aggressive behaviours at school. The content and specific focus of such programs, however, may vary depending upon participants' gender and other consideration

Furlong, M. J., You, S., Smith, D. C., Gonzlaez, V., Boman, P. (2012). An Examination of the Factorial Invariance of the Multidimensional School Anger Inventory for Five Pacific Rim Countries

This investigation evaluated the psychometric properties of the Multidimensional School Anger Inventory (MSAI) using samples from the five Pacific Rim countries of Australia, Guatemala, Japan, Peru, and the United States (USA). The combined sample included 3,181 adolescents (M = 14.8 years) with 52% females and was used to examine factorial invariance across the five national X gender groups. The analysis was completed in two stages. First, CFA examined configural invariance for the four MSAI factors: Anger Experience (AE), Hostility (Ho), Destructive Expression (DE), and Coping (Cop). This four-factor model did not converge for the Peruvian students.  Second, with the other half of the split sample and using the top four loaded items for AE, Ho, and DE, which form the core of Spielberger’s Anger-Hostility-Aggression (A-H-A) model, a CFA found configural invariance, partial metric invariance, and partial scalar invariance. Latent means analysis for AE, HO, and DE showed that compared to the USA sample, the other students showed higher levels of AE (ds = .37 to .73). In addition, the Australian (d = .40) and Japanese students (d = .21) had significantly higher scores on Ho. Australian students also had higher scores on DE (d = .30), but the Japanese students had lower scores (d = -.17). Comparisons of latent mean scores by gender found no differences between males and females in any country in reported AE. The largest gender differences were for DE among Australian (d = -.67), Guatemalan (d = -.42), and the USA (d = -.66) students. This study supported the factorial invariance of an abbreviated MSAI that can be efficiently used in cross-national studies; however, researchers should attend to possible mean differences across countries. Implications for the use of the MSAI in comparative research are discussed.    


© MSAI (Smith & Furlong) 2012