Development and Validation of a Short Version of the Primary Scales of the Inventory of Personality Organization: A Study among Japanese University Students

Abstract

Background: The purpose of this study was to develop a short version of the Inventory of Personality Organization (IPO) among Japanese university students. The measurement invariance between men and women was investigated. Construct validity was also examined. Method: Student data from two universities in Japan (N = 504) were used. We used the main domains of the IPO: Primitive Defences, Identity Diffusion, and Reality Testing. Three items with the highest item-total correlation were extracted from those belonging to each domain. EFAs and CFAs compared different factor structure models. The best model’s measurement invariance was examined. Results: The subscales scores of the short version IPO were correlated with the other variables that were theoretically expected to be associated. Results indicated that a 9-item 3-factor model was the best fit (χ2/df = 1.768, CFI = 0.970). Internal consistency of each domain was good. Between men and women, this model showed stability as factor covariance invariance. The correlation between the full and short versions was above 0.8 for each subscale. The associations of the three subscales with the other variables were in the expected direction. Conclusion: The 9-item version of the IPO may be useful in clinical and research settings.