ABSTRACT
The "Questionnaire of Thoughts and Feelings" (QTF) is being used as screening instrument as well as tool for treatment planning and treatment evaluation for patients with borderline personality disorder (BPD). The primary goal of this study was the validation of the dimensional structure of the short version with 14 items, QTF-14. Additionally, item characteristics, reliability and evidence of convergent validity were examined. A diagnostically homogenous sample of patients with BPD (N=3035) of a psychosomatic clinic was presented with several self-assessment inventories, including the QTF-14. The expected single-factor model of the structure of the FGG showed unacceptable model fit indices (CFI=0.751; TLI=0.706; RMSEA=0.115; SRMR=0.073). Exploratory factor analysis showed evidence of two or three specific factors. In a subsequent confirmatory factor analysis, a bifactor model with two specific factors proved to be preferable (CFI=0.956; TLI=0.936; RMSEA=0.054; SRMR=0.034). The internal consistency of the total scale as well as the suggested subscales "Relationship Difficulties and Emotional Dysregulation" and "Autoaggression" was acceptable to good (ω=.81-.84; α=.79-.85). Associations with similar scales were as expected. Good psychometric properties of the QTF-14 can be confirmed in this study. Using the suggested subscales could support treatment planning.
Subject(s)
Borderline Personality Disorder , Humans , Borderline Personality Disorder/diagnosis , Reproducibility of Results , Surveys and Questionnaires , Emotions , PsychometricsABSTRACT
There is a growing consensus that emotion dysregulation is the defining feature of borderline personality disorder, but there are also persisting competing views on the pathology and dimensional structure of the disorder. The "Impulsivity and Emotion Dysregulation Scale" (IED-27) is already being used for assessing patients with borderline personality disorder. The particular aim of this study was to gain evidence of the dimensional structure of the disorder by confirmatory factorial analysis. A heterogeneous sample of patients with borderline personality disorder (n=484), attention deficit (hyperactivity) disorder (n=191) and eating disorders (n=197) of a psychosomatic clinic were given general and specific self-assessment inventories via computer-assisted testing system, including the IED-27. Results showed a bifactor model with a general factor and three specific factors: "emotional dysregulation", "relationship difficulties" and "self-injuries and suicidal behaviours", to be the model with the best fit (ΔCFI=0,032). The internal consistency of the sample seems to be good (ω=0,83-0,93) and expected correlations to similar subscales could be proved. Good psychometric criteria of the IED-27 were replicated in this study. Prospectively, the subscales allow a more detailed representation and specific treatment planning.
Subject(s)
Attention Deficit Disorder with Hyperactivity , Borderline Personality Disorder , Borderline Personality Disorder/diagnosis , Emotions , Humans , Impulsive Behavior , PsychometricsABSTRACT
Borderline personality disorder (BPD) is a severe mental disorder which is accompanied by impaired functioning, complex psychosocial problems and incurs large costs. However, it rarely has been examined whether BPD symptoms form a hierarchical (Mokken) scale, which would imply that symptoms are ordered along a latent dimension of borderline severity. In this study the data from 1,198 inpatients of a psychosomatic clinic were examined. A non-parametric Mokken analysis was employed to investigate if the BPD symptoms measured by the impulsivity and emotion dysregulation scale (IES-27) formed a hierarchical scale. 16 of the 27 items, mainly regarding emotion dysregulation, formed a monotonous Mokken scale with a very strong hierarchy of BPD symptoms. These results imply that single BPD symptoms provide information about BPD severity, suggesting several important consequences regarding escalation of BPD, diagnosis and treatment, e.g. how to optimize treatment planning.
