Related Experiment Videos
A note on the selection of control groups and control variables in comorbidity research
1Obesity Research Center, Saint Luke's/Roosevelt Hospital, Columbia University College of Physicians and Surgeons, New York, NY 10025.
Comprehensive Psychiatry
|September 1, 1993
Summary
Investigating comorbidity in eating disorders (EDs) requires testing two hypotheses. Proper control groups are essential for accurately demonstrating that ED patients exhibit more psychopathology than chance would predict.
Area of Science:
- Psychiatry
- Clinical Psychology
- Eating Disorder Research
Background:
- Comorbidity in eating disorders (EDs) is a growing area of research interest.
- Accurate assessment of comorbidity is crucial for understanding and treating EDs.
Purpose of the Study:
- To highlight the necessity of rejecting two specific null hypotheses when demonstrating comorbidity in EDs.
- To emphasize the importance of selecting appropriate control groups for comorbidity research.
Main Methods:
- The study illustrates the concepts using simulated data.
- It outlines the requirements for testing null hypotheses related to comorbidity.
Main Results:
- Demonstrating comorbidity requires rejecting the null hypothesis that ED patients have no more symptomatology than controls (H0-1).
- It also necessitates rejecting the null hypothesis that ED patients do not have more specific psychopathology than other non-ED-specific psychopathology (H0-2).
Conclusions:
- The choice of control group is critical for testing H0-1 in ED comorbidity studies.
- Testing H0-2 is essential for valid comorbidity assessment in EDs and other conditions.
- The principles discussed apply broadly to comorbidity research, with adjustments for specific hypotheses and comparison groups.