Rewards of bridging the divide between measurement and clinical theory: demonstration of a bifactor model for the Brief Symptom Inventory.

Department of Psychology, Arizona State University, 950 S. McAllister, P.O. Box 871104, Tempe, AZ 85287, USA.
Psychological Assessment (Impact Factor: 2.99). 07/2011; 24(1):101-13. DOI: 10.1037/a0024712
Source: PubMed

ABSTRACT There is growing evidence that psychiatric disorders maintain hierarchical associations where general and domain-specific factors play prominent roles (see D. Watson, 2005). Standard, unidimensional measurement models can fail to capture the meaningful nuances of such complex latent variable structures. The present study examined the ability of the multidimensional item response theory bifactor model (see R. D. Gibbons & D. R. Hedeker, 1992) to improve construct validity by serving as a bridge between measurement and clinical theories. Archival data consisting of 688 outpatients' psychiatric diagnoses and item-level responses to the Brief Symptom Inventory (BSI; L. R. Derogatis, 1993) were extracted from files at a university mental health clinic. The bifactor model demonstrated superior fit for the internal structure of the BSI and improved overall diagnostic accuracy in the sample (73%) compared with unidimensional (61%) and oblique simple structure (65%) models. Consistent with clinical theory, multiple sources of item variance were drawn from individual test items. Test developers and clinical researchers are encouraged to consider model-based measurement in the assessment of psychiatric distress.

  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Knowledge of coping styles is useful in clinical diagnosis and suggesting specific therapeutic interventions. However, the latent structures and relationships between different aspects of coping styles have not been fully clarified. A full information item bifactor model will be beneficial to future research. One goal of this study is identification of the best fit statistical model of coping styles. A second goal is entails extended analyses of latent relationships among different coping styles. In general, such research should offer greater understanding of the mechanisms of coping styles and provide insights into coping with stress. Coping Styles Questionnaire (CSQ) and Generalized Self-Efficacy Scale (GSES) were administrated to officers suffering from military stress. Confirmatory Factor Analyses was performed to indentify the best fit model. A hierarchical item response model (bifactor model) was adopted to analyze the data. Additionally, correlations among coping styles and self-efficacy were compared using both original and bifactor models. Results showed a bifactor model best fit the data. Item loadings on general and specific factors varied among different coping styles. All items loaded significantly on the general factor, and most items also had moderate to large loadings on specific factors. The correlation between coping styles and self-efficacy and the correlation among different coping styles changed significantly after extracting the general factor of coping stress using bifactor analysis. This was seen in changes from positive (r = 0.714, p<0.01) correlation to negative (r = -0.335, p<0.01) and also from negative (r = -0.296, p<0.01) to positive (r = 0.331, p<0.01). Our results reveal that coping styles have a bifactor structure. They also provide direct evidence of coexisting coping resources and styles. This further clarifies that dimensions of coping styles should include coping resources and specific coping styles. This finding has implications for measurement of coping mechanisms, health maintenance, and stress reduction.
    PLoS ONE 01/2014; 9(5):e96451. · 3.53 Impact Factor
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: The Derogatis symptom checklist (SCL-90-R) and its short version, the Brief Symptom Inventory (BSI), are widely used instruments, despite the fact that their factor structures were not clearly confirmed. The goals of this research were to compare four measurement models of these instruments including one-factor, nine-factor, a second-ordered factor model and a bifactor model, in addition to testing the gender difference in symptom factors in a community sample. SCL-90-R was assessed in a large community survey which included 2,710 adults who represent the population of Hungary. Statistical analyses included a series of confirmatory factor analyses and multiple indicator multiple cause (MIMIC modeling). The responses to items were treated as ordinal scales. The analysis revealed that the bifactor model yielded the closest fit in both the full SCL-90-R and BSI, however the nine-factor model also had an acceptable level of fit. As for the gender differences, women scored higher on global severity, somatization, obsession-compulsion, interpersonal sensitivity, depression and anxiety factors. Men scored higher on hostility and psychoticism. The bifactor model of symptom checklist supports the concept of global symptom severity and specific symptom factors. Global symptom severity explains the large correlations between symptom factors.
    Psychiatry Research. 01/2014;
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Research on comorbidity of psychiatric disorders identifies broad superordinate dimensions as underlying structure of psychopathology. While a syndrome-level approach informs diagnostic systems, a symptom-level approach is more likely to represent the dimensional components within existing diagnostic categories. It may capture general emotional, cognitive or physiological processes as underlying liabilities of different disorders and thus further develop dimensional-spectrum models of psychopathology. Exploratory and confirmatory factor analyses were used to examine the structure of psychopathological symptoms assessed with the Brief Symptom Inventory in two outpatient samples (n=3171), including several correlated-factors and bifactor models. The preferred models were correlated with DSM-diagnoses. A model containing eight correlated factors for depressed mood, phobic fear, aggression, suicidal ideation, nervous tension, somatic symptoms, information processing deficits, and interpersonal insecurity, as well a bifactor model fit the data best. Distinct patterns of correlations with DSM-diagnoses identified a) distress-related disorders, i.e., mood disorders, PTSD, and personality disorders, which were associated with all correlated factors as well as the underlying general distress factor; b) anxiety disorders with more specific patterns of correlations; and c) disorders defined by behavioural or somatic dysfunctions, which were characterised by non-significant or negative correlations with most factors. This study identified emotional, somatic, cognitive, and interpersonal components of psychopathology as transdiagnostic psychopathological liabilities. These components can contribute to a more accurate description and taxonomy of psychopathology, may serve as phenotypic constructs for further aetiological research, and can inform the development of tailored general and specific interventions to treat mental disorders.
    Comprehensive psychiatry 11/2013; · 2.08 Impact Factor