Related Experiment Video
Updated: Mar 22, 2026
05:21
Computerized Adaptive Testing System of Functional Assessment of Stroke
Published on: January 7, 2019
6.4K
Test-Retest Reliability of a Computerized Adaptive Depression Screener
David Beiser1, Milkie Vu1, Robert Gibbons1
1Dr. Beiser and Ms. Vu are with the Section of Emergency Medicine, and Dr. Gibbons is with the Center for Health Statistics, University of Chicago, Chicago. Send correspondence to Dr. Gibbons (e-mail: rdg@uchicago.edu ).
Psychiatric Services (Washington, D.C.)
|April 16, 2016
Summary
Computerized adaptive testing (CAT) offers precise depression screening with reduced burden. The CAT Depression Inventory (CAT-DI) demonstrated excellent test-retest reliability in emergency department patients, validating its use.
Area of Science:
- Psychological assessment
- Clinical psychology
- Medical informatics
Background:
- Computerized adaptive testing (CAT) enhances measurement precision and reduces patient burden compared to traditional tests.
- Concerns exist regarding the reliability of CAT due to variable item administration between and within individuals over time.
- The CAT Depression Inventory (CAT-DI) is used for depression assessment, particularly in screening settings with predominantly normal scores.
Purpose of the Study:
- To measure the test-retest reliability of the CAT Depression Inventory (CAT-DI).
- To assess the reliability of CAT-based depression screening in an academic emergency department (ED) setting.
Main Methods:
- A random sample of 101 adults in an academic ED were screened twice with the CAT-DI during their visit.
- Test-retest scores, bias, and reliability were statistically assessed.
Main Results:
- Fourteen percent of patients scored in the mild depression range, 4% moderate, and 3% severe.
- Test-retest scores showed no significant bias.
- Excellent test-retest reliability was observed (r=.92).
Conclusions:
- The CAT-DI provides reliable depression screening results for emergency department patients.
- Concerns regarding the impact of varying item presentation on test-retest reliability were not supported by the findings.
