Abstract
Our associated paper presented a psychometric evaluation of the Wender Utah Rating Scale (WURS) and its abbreviated version, the WURS-25. Instead of actual factors scores, we employed “item averages” calculated by the average score of each item comprising that factor. We did not present a factor analysis of the WURS-25. Herein we identify items of the full WURS that are redundant or not part of any of the scale's five factors. Removing these items produced a shortened version, the WURS-45. We performed a logistic regression using actual factor loadings as well as factors based on item averages, and compared major depressive disorder (MDD) to generalized anxiety disorder (GAD) patients in the same analysis. We performed exploratory factor analysis with the WURS-45 items. We then performed logistic regressions and Receiver Operating Characteristics (ROC) analyses with the WURS-45 and WURS-25 factors. No increase in specificity or sensitivity arose when actual factors scores were used as opposed to factor scores from item averages. MDD and GAD ROC curves were very similar, supporting combining MDD with GAD patients into a single group. WURS-45 factors paralleled those derived from the full WURS. ROC curves, logistic regression and confusion tables showed the WURS-45 preserved the excellent diagnostic separation produced by the full WURS. Similar analyses showed WURS-25 scoring using its three factors improved its diagnostic utility. The WURS-45 has reduced redundancy with minimal loss in discriminatory power. Analysis of the WURS-25 using factor scores boosts its performance. Both versions of the scale provide clinical information describing childhood ADHD and are useful in separating adult patients with ADHD from those with MDD or GAD.
Original language | English (US) |
---|---|
Article number | 107232 |
Journal | Data in Brief |
Volume | 37 |
DOIs | |
State | Published - Aug 2021 |
All Science Journal Classification (ASJC) codes
- General