Related Experiment Videos
The Goldilocks dilemma in survey design and its solution
1Department of Statistics, University of Wisconsin, Madison 53706.
Journal of Clinical Epidemiology
|March 1, 1993
Summary
Designing surveys with followups requires careful timing to avoid missing event occurrences. This study presents a technique to optimize followup intervals in multiwave studies, preventing inefficient data collection for events like functional dependence in the elderly.
Area of Science:
- Gerontology
- Survey Methodology
- Biostatistics
Background:
- Designing surveys with followups to detect event occurrences presents a timing dilemma.
- Followup intervals that are too short risk observing few events, while intervals too long risk observing nearly all events, both being uninformative.
- This challenge is amplified in multiwave studies with multiple followups.
Purpose of the Study:
- To address the 'Goldilocks dilemma' in survey design concerning optimal followup intervals.
- To introduce a technique for guiding the spacing of followup waves in multiwave studies.
- To prevent inefficiencies in study design related to event detection.
Main Methods:
- Development of a novel technique for determining optimal followup wave spacing.
- Application of the technique to data from the Massachusetts Health Care Panel Study (MHCPS).
- Demonstration using data on functional dependence in the elderly.
Main Results:
- The described technique provides guidance for spacing followup waves in multiwave studies.
- This method helps mitigate the risk of uninformative data due to suboptimal timing.
- The Massachusetts Health Care Panel Study data serves as a practical example for the technique's application.
Conclusions:
- The proposed technique offers a partial solution to the challenge of timing followups in event-detection surveys.
- Optimizing followup intervals is crucial for efficient study design and reliable data collection.
- This approach is particularly relevant for longitudinal studies investigating the onset of events in specific populations, such as the elderly.