Search

Former studies have repeatedly found the psychoticism (PSY) scale of Symptom Checklist-90-Revised to be a heterogeneous construct. The aim of this study was to confirm and further explore the nature of this heterogeneity within a large sample of patients with mainly personality disorders. Within a total sample of 3 794 patients, one-half was randomly selected for explorative factor analysis in order to assess the internal structure of the PSY scale and the other half to cross-validate the findings by a confirmatory factor analysis. The total sample was then used to assess associations between the components from the factor analyses and several clinical measures and diagnoses. A one-factor solution of the PSY scale yielded poor fit to the data, but a proposed structure of three latent constructs was confirmed by good model fit. The three subsets of the PSY scale, labelled metacognitive dysfunction, self-accusation and detachment, shared variance with different personality disorders and different aspects of psychopathology, e.g. previous psychotic episodes. The heterogeneous PSY scale of SCL-90-R can be divided into three meaningful clinical concepts, reflecting different aspects of psychosis-near experiences. The factors warrant confirmation in other populations. Copyright © 2014 John Wiley & Sons, Ltd., (Copyright © 2014 John Wiley & Sons, Ltd.)

Patients with personality disorders represent maladaptive attachment strategies, influencing the quality of their interpersonal relationships. Within these patient populations, it is important to have a measure, easily applied, to assess such strategies. A widely used instrument is Experiences in Close Relationships (ECR), assessing two constructs called avoidance and anxiety. Lately, two short forms of ECR have been proposed, called ECR-S and ECR-N12. The aim of this study is to investigate the psychometric properties of the Norwegian @version of ECR and its two short forms in a sample of patients with mainly personality disorders. An internal consistency and confirmatory factor analysis of ECR, ECR-S and ECR-N12 from 495 patients was conducted, as well as an exploratory factor analysis of ECR. The internal consistency of ECR was found questionable. Confirmatory factor analysis revealed a poor model fit based on ECR and ECR-S. The ECR-N12 revealed a mediocre fit, indicating a potential for improvement. Exploratory factor analysis indicated two different aspects of avoidance and three aspects of anxiety. This five-factor solution was called ECR-FF. Inferences from scores based on ECR should be derived with care. A revision of ECR and ECR-N12 is warranted, and further studies are needed to investigate the validity of ECR-FF., (Copyright © 2015 John Wiley & Sons, Ltd.)