Rorschach Assessment of Traumatized Refugees: An Exploratory Factor Analysis

Abstract

Fifty-one multitraumatized mental health patients with refugee backgrounds completed the Rorschach (Meyer & Viglione, 2008), Harvard Trauma Questionnaire, and Hopkins Symptom Checklist–25 (Mollica, McDonald, Massagli, & Silove, 2004), and the World Health Organization Quality of Life-BREF questionnaire (WHOQOL Group, 1998) before the start of treatment. The purpose was to gain more in-depth knowledge of an understudied patient group and to provide a prospective basis for later analyses of treatment outcome. Factor analysis of trauma-related Rorschach variables gave 2 components explaining 60% of the variance; the first was interpreted as trauma-related flooding versus constriction and the second as adequate versus impaired reality testing. Component 1 correlated positively with self-reported reexperiencing symptoms of posttraumatic stress (r = .32, p < .05). Component 2 correlated positively with self-reported quality of life in the physical, psychological, and social relationships domains (r = .34, .32, and .35, p < .05), and negatively with anxiety (r = –.33, p < .05). Each component also correlated significantly with resources like work experience, education, and language skills.

Acknowledgments

This research was fully financed by the Norwegian Centre for Traumatic Stress Studies, Oslo, Norway. We would like to thank Sverre Varvin for adapting the frames of the research program for this study, Thorleif Lund and Tore Wentzel-Larsen for statistical advice, and Jim Allen and Beate Øhre for helpful comments.

Marianne Opaas presented preliminary findings from the study at the Society of Personality Assessment annual meeting, San José, California, 2010 and at the Society of Psychotherapy Research annual meeting in Bern, Switzerland, 2011.

Notes

We are in debt to Jim Allen for alerting us in an early phase of our study to the concept of the biphasic response to trauma, to the Rorschach findings of traumatized samples, and to the possibility of using factor analysis to explore this.

The criteria defined for a behaviorally defensive and constricted response-character style was R < 21 and Lambda > 0.55; for an exaggerated or dilated style it was R > 21 and Lambda < 0.55. (Lambda is interpretatively similar to F%, indicating a tendency to simplify complex stimuli fields; Meyer et al., 2001).