Related Experiment Videos
Measuring aggression in older adults: a latent variable modeling approach.
K J O'Malley1, C A Orengo, M E Kunik
1Houston Center for Quality of Care and Utilization Studies, Baylor College of Medicine, USA. komalley@bcm.tmc.edu
Aging & Mental Health
|September 10, 2002
Summary
This study developed a new method using latent variable modeling to combine aggression scales into a single, reliable score for dementia patients. This approach offers researchers a more comprehensive way to measure aggression.
Area of Science:
- Psychiatry
- Gerontology
- Psychometrics
Background:
- Aggression is a common symptom in dementia patients.
- Existing methods for measuring aggression often lack comprehensive reliability.
- There is a need for improved measurement tools in geriatric psychiatry.
Purpose of the Study:
- To demonstrate a technique for combining multiple aggression scales into a single score.
- To utilize latent variable modeling for enhanced aggression measurement.
- To improve the reliability and comprehensiveness of aggression assessment in dementia.
Main Methods:
- A cross-sectional measurement study design.
- Latent variable modeling to integrate data from seven aggression scales.
- Analysis of data from 49 male dementia patients at a Veterans Affairs Medical Center.
Main Results:
- A unidimensional latent variable model of aggression adequately represented the data.
- The reliability of the combined aggression score was estimated at 0.90.
- Individual aggression scales had reliabilities less than 0.84 in this sample.
Conclusions:
- Latent variable modeling provides a reliable method for creating a composite aggression score.
- This technique offers advantages over traditional methods for measuring aggression in research.
- The findings support the use of this combined score for comprehensive aggression assessment in dementia.