Detecting and Explaining Aberrant Responding to the Outcome Questionnaire-45

Assessment. 2015 Aug;22(4):513-24. doi: 10.1177/1073191114560882. Epub 2014 Dec 16.

Abstract

We applied item response theory based person-fit analysis (PFA) to data of the Outcome Questionnaire-45 (OQ-45) to investigate the prevalence and causes of aberrant responding in a sample of Dutch clinical outpatients. The [Formula: see text] person-fit statistic was used to detect misfitting item-score patterns and the standardized residual statistic for identifying the source of the misfit in the item-score patterns identified as misfitting. Logistic regression analysis was used to predict person misfit from clinical diagnosis, OQ-45 total score, and Global Assessment of Functioning code. The [Formula: see text] statistic classified 12.6% of the item-score patterns as misfitting. Person misfit was positively related to the severity of psychological distress. Furthermore, patients with psychotic disorders, somatoform disorders, or substance-related disorders more likely showed misfit than the baseline group of patients with mood and anxiety disorders. The results suggest that general outcome measures such as the OQ-45 are not equally appropriate for patients with different disorders. Our study emphasizes the importance of person-misfit detection in clinical practice.

Keywords: Outcome Questionnaire–45; aberrant responding; item response theory; outcome measurement; person-fit analysis.

Publication types

  • Research Support, Non-U.S. Gov't

MeSH terms

  • Adolescent
  • Adult
  • Aged
  • Bias*
  • Female
  • Humans
  • Logistic Models
  • Male
  • Middle Aged
  • Outcome and Process Assessment, Health Care
  • Surveys and Questionnaires