Abstract

In the present study, the validity scales of the Personality Assessment Inventory (PAI) of the true victim group and the fake victim group were used to determine the cutoff points for discriminating between the two groups of adults in their twenties. PAI scores of true crime victims who participated in a victim assessment system in Korea (true victim group, n = 99, approximately 90% of females) were provided by the National Police Agency, and PAI scores were collected from community members who had never been victimized. Some of the community members were instructed to fake victimization (fake victim group, n = 65, approximately 60% of females), and the others were instructed to answer honestly about their current status (normal group, n = 54, approximately 55% of females). Results showed that the fake victim group had statistically significantly higher T-scores on the infrequency (INF) and negative impression (NIM) scales than the true victim group. For the INF, a T-score of 60-66 had the highest accuracy and lowest false negative rate. Using 61 T-score points on the NIM scale as the cutoff point provided the most accurate discrimination between true and fake victims, with an accuracy rate of .69. In addition, using 84 T-score points on the NIM scale as the cutoff point, which can minimize the false negative rate, reduced the false negative rate to 8%, and the accuracy rate was .64. Compared with the normal group, the PAI clinical scale profiles of the true and fake victim groups were similar, but the fake victim group scored approximately 4 to 20 T-score points higher than the true victim group on the scales of physical appeal (SOM), paranoia (PAR), schizophrenia (SCZ), antisocial traits (ANT), and alcohol problems (ALC). The implications for the use of the PAI in the Korean Victim Assessment System were discussed.