There are two versions of the MSAI. Both versions have sound psychometric properties that support their use in research and for evaluating school-based mental health services. We offer the following suggestions to guide your thinking about which version best fits your interests.


This MSAI version has strong psychometric properties and its core factor structure has been replicated in four studies using samples of students from different countries. It also has the advantage of including the Coping subscale, which is an important consideration when evaluating a youth's anger-related experiences.  Given its more comprehensive coverage, the 36-item MSAI is recommended for use in the evaluation of school-based anger management interventions and as a counseling assessment resource. In such circumstances, there is less time demand and the primary interest is not just to efficiently measure the MSAI latent traits, as is more true in model testing research, but in developing a clearer, more comprehensive understanding of a specific youth's anger management concerns.


The most comprehensive study of the MSAI is the mtulitgroup factor analysis by Furlong et al. (2012), which showed that:

  • When the factor structure of the 36-item MSAI was compared across five Pac Rim countries, the construct validity of the four subscales (Anger Expression, Hostility, Destructive Expression, and Coping) were not retained. 
  • Although the four-factor structure held for Australian, Japan, Guatemala, and the United States, it did not hold for Peru. 
  • In the interest of identifying a version of the MSAI that holds the best promise for use in comparative cross-national research and to provide a briefer measure for all researchers, we carried out additional analyses. We found that  a subset of 12 MSAI items (4 each of Anger Experience, Hostility, and Destructive Expression) had a good fit with the prevalent anger-hostility-aggression model. And, the A-H-A model demonstrated factorial  invariance across all five countries.
  • The psychometric properties of the MSAI-12 are strong, theoretically based, and have been validated across diverse samples of adolescents. For this reason, we recommend the use of this scale in most research projects. It also has the advantages of lessening the response burden on study participants and it can be used efficiently in studies that examine more complex structural equation models. 

Gender Considerations

In previous MSAI research, females scored higher on Coping and lower on Hostility than males; whereas males had higher scores on Destructive Expression (Furlong et al., 2002). In another study, Australian boys and girls had no differences in the intensity with which they experienced anger (Boman, 2003), but in a sample of Iranian students, just the opposite was found (Ghanizadeh, 2008); however, both of these studies reported that girls had lower levels of cognitive hostility than boys (Bowman & Yates, 2001). An interesting finding of the Boman et al. (2006) study was that for females, higher levels of anger did not always result in more aggressive response tendencies; that is, females felt anger as often as males, but were less likely to endorse the use of aggressive responses.    

The Furlong et al. (2012) study examined gender differences within five Pac Rim counties and found consistent differences on the Destructive  Expression subscale (ES = - .31 to - .67) with females having lower mean scores. Only for the USA and Japanese samples did the males have higher Hostility scores than females (ES = -.35 and -.45). The USA sample was the only one with the females reporting less intense levels of Anger Experience than males (ES = -.32). Download table showing gender and cross-national latent mean comparisons.

Age Considerations

The MSAI has been used with children as young as 10 years-old. However, the bulk of validation research has been with adolescents ages 12-17. 

© MSAI (Smith & Furlong) 2012