Examining the Factor Structure of the Moral Injury Events Scale in a Veteran Sample

• Published in: Military medicine Vol. 185; no. 1-2; pp. e75 - e83
• Main Authors: Richardson, Cameron B, Chesnut, Ryan P, Morgan, Nicole R, Bleser, Julia A, Perkins, Daniel F, Vogt, Dawne, Copeland, Laurel A, Finley, Erin
• Format: Journal Article
• Language: English
• Published: England Oxford University Press 13.02.2020
• Subjects: Humans; Military Personnel; Moral injury; Morals; Stress Disorders, Post-Traumatic; Surveys and Questionnaires; Veterans; Potentially Morally Injurious Experiences; Moral Pain; Moral Injury Events Scale; Factor Analysis; Veteran; Moral Injury
• ISSN: 0026-4075; 1930-613X; 1930-613X
• DOI: 10.1093/milmed/usz129

Abstract Introduction Interest in moral injury has burgeoned over the last decade as an increasing number of professionals recognize that current conceptualizations of trauma are not sufficient to explain some of the challenges that military service members face. The Moral Injury Events Scale (MIES) was the first instrument developed to measure exposure to military events that could produce moral injury (Nash et al., 2013). Two previously published validation studies that focused on service members still in uniform reveal some discrepancies regarding the scale’s factor structure. Bryan and colleagues (2016) documented a three-factor solution (i.e., transgressions-others, transgressions-self, and betrayal), while Nash and colleagues (2013) noted a two-factor solution (i.e., perceived transgressions and perceived betrayals). Materials and Methods With an interest in further scale validation, the present study utilized survey data from a group of veterans recently separated from service, to test the fitness of the two- or three-factor model and to examine the invariance of the scale across gender and branch. The investigators followed the policies governing the protection of human subjects as prescribed by ICF Institutional Review Board. Results Findings revealed a two-factor structure that differed from previous factor solutions (transgressions-self and transgressions-others) and was invariant across gender and branch. High intercorrelations among the MIES items that addressed exposure and reaction to events suggested that these experiences tend to co-occur. Removal of the event items did not significantly impact model fit. Conclusions The factor structure identified in the present study aligns with current theoretical conceptualizations of moral injury. The inability to distinguish between event and reaction items and the lack of impact on the factor structure when event items were removed from the model suggest the MIES is more accurately described as a measure of moral pain as compared to a measure of potentially morally injurious experiences (PMIEs). Given that all events associated with moral injury have impinged upon welfare, justice, rights, and fairness considerations, one could argue that little would be gained by focusing attention on the development of a measure of PMIEs. Given the lack of convergence in factor structure across samples, future directions may profit from a focus on further clarifying the temporal stability of the MIES factor structure.