Journal or Personality Disorders. 14(3), 189-198,2000 2000 The Guilford Press MODELING AND MEASURING THE PERSONALITY DISORDERS Gordon Parker, DSc, MD, PhD, Dusan Hadzi-Pavlovic, BSc, MPsychol, and Kay Wilhelm, MD Livesley, Jang, and Vernon (1998. Archives of General Psychia try, 55, 941-948) reported that personality disorders (PDs) are quantitatively extreme expressions of normal personality func tioning. A similarly designed study attempts to replicate those findings for both self- and observer-rated reports of patients judged clinically to have a PD. Analyses of data sets generated by 758 self- reports (SRs) and 5 1 5 reports from corroborative witnesses (CWs) refined the set of 266 descriptors to 142 items assessing 30 constructs. Intercorrelation of the constructs re vealed considerable interdependence. Principal components analyses identified four factors, consistent across the SR and CW databases, and consistent with the Livesley et al. (1998) analyses where they were labeled Emotional Dysregulation, Dissocial, Inhibition, and Compulsivity. Replication and exten sion of findings to CW-rated data offer additional support for the argument that higher-order PD traits strongly resemble normal personality dimensions. While there are multiple models for conceptualizing and measuring the per sonality disorders (PDs), the keenest debate is over categorical versus di mensional approaches. Previous and current Diagnostic and Statistical Manual of Mental Disorders (DSM; American Psychiatric Association, 1980, 1987, 1994) and International Classification of Diseases (ICD-10; World Health Organization, 1992) classifications impute a number of distinct, in dependent and categorical PDs and, at least in the last three DSM editions, three higher-order PD "clusters." Such a categorical approach reflects the historical preference by psychiatrists to assign patients to classes, or at least to diagnostic groups, rather than adopting a dimensional approach From the School of Psychiatry and Mood Disorders Unit, University of New South Wales, Sydney. This research is based on data collected previously by Parker et al. ( 1 996) and was supported by an NHMRC Program Grant (993208), and an Infrastructure Program Grant from the NSW Department of Health. The assistance of Elaine Barrett. Kerrie Eyers. Gemma Gladstone, Penelope Irvine, Heather Brotchie, Christine Taylor, Yvonne Foy, and the many psychiatrists who contributed data is gratefully acknowledged. Address correspondence to Gordon Parker, Psychiatry Unit, Prince of Wales Hospital, Rand- wick 2031, Australia. 189