Brief Psychiatric Rating Scale Expanded Version: How do new items affect factor structure?
| Autor: | A. John Rush, Stephen Shon, Trisha Suppes, Madhukar H. Trivedi, Alexander L. Miller, Dawn I. Velligan, Marcia G. Toprac, Kathy Shores-Wilson, Ellen B. Dennehy, John A. Chiles, Bradley Witte, M. Lynn Crismon, Melanie M. Biggs, Thomas J. Carmody, Thomas J. Prihoda, T. Michael Kashner |
|---|---|
| Rok vydání: | 2005 |
| Předmět: |
Adult
Male Texas Medication Algorithm Project medicine.medical_specialty Psychometrics Mental Disorders Test validity Severity of Illness Index Confirmatory factor analysis Exploratory factor analysis Psychiatry and Mental health Goodness of fit Brief Psychiatric Rating Scale medicine Humans Anxiety Female medicine.symptom Factor Analysis Statistical Psychiatry Psychology Algorithms Biological Psychiatry |
| Zdroj: | Psychiatry Research. 135:217-228 |
| ISSN: | 0165-1781 |
| DOI: | 10.1016/j.psychres.2005.05.001 |
| Popis: | Our goal was to suggest a factor structure for the Brief Psychiatric Rating Scale Expanded Version (BPRS-E) based upon a large and diverse sample and to determine which of the new items improved the factors derived from the 18-item version of the scale that have been used in clinical research for decades. We investigated the consistency of our proposed model over time and across demographic groups. As part of the Texas Medication Algorithm Project, the BPRS-E was administered to a total of 1440 psychiatric outpatients in three different diagnostic groups on multiple occasions. The sample was randomly split so that exploratory factor analysis could be done with the first half, and the model could be confirmed on the second half. A four-factor structure including factors assessing depression/anxiety, psychosis, negative symptoms, and activation was found. For each factor, we specify items in the expanded version that added to the breadth of the commonly used clinical factors while improving or maintaining goodness of fit and reliability. The final model proposed was consistent over time and across diagnosis, phase of illness, age, gender, ethnicity, and level of education. The BPRS-E has a stable four-factor structure, making it useful as a clinical outcome measure. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|
Full Text from ScienceDirect