Related Experiment Videos
Score comparability of short forms and computerized adaptive testing: Simulation study with the activity measure for
Stephen M Haley1, Wendy J Coster, Patricia L Andres
1Research and Training Center on Measuring Rehabilitation Outcomes, Center for Rehabilitation Effectiveness, Sargent College of Health and Rehabilitation Sciences, Boston University, Boston, MA, USA. smhaley@bu.edu
Archives of Physical Medicine and Rehabilitation
|April 15, 2004
Summary
Computerized adaptive testing (CAT) and short forms provide accurate scoring estimates for the Activity Measure for Post-Acute Care (AM-PAC) domains. CAT offers potential advantages in efficiency and precision for individual scoring.
Area of Science:
- Rehabilitation Medicine
- Health Outcomes Measurement
- Psychometrics
Background:
- The Activity Measure for Post-Acute Care (AM-PAC) is crucial for assessing patient function in various post-acute settings.
- Evaluating the accuracy of abbreviated assessment tools is essential for efficient and effective patient management.
Purpose of the Study:
- To compare the scoring accuracy of simulated short-form instruments and computerized adaptive testing (CAT) against comprehensive item sets for the AM-PAC.
- To determine the reliability and validity of these abbreviated methods across three key functional domains.
Main Methods:
- A prospective study was conducted across six post-acute care networks involving 485 adult volunteers.
- Short-form and CAT applications were developed for physical & mobility, personal care & instrumental, and applied cognition domains using existing and new items.
- Simulated scores from short forms and CAT were compared to scores derived from complete AM-PAC item pools.
Main Results:
- Simulated CAT scores demonstrated high correlations with total item pool scores across all domains (r range, .90–.98).
- 10-item short forms provided good estimates for physical & mobility and personal care & instrumental domains but were less consistent for applied cognition.
- Confidence intervals for individual scores were narrower with CAT compared to short forms.
Conclusions:
- Both setting-specific short forms and CAT provide accurate scoring estimates for AM-PAC domains.
- CAT's adaptive item selection enhances its ability to match individual responses, potentially offering greater precision.
- CAT may present advantages in practicality, efficiency, and individual scoring precision over traditional short forms.