Related Experiment Videos
A power analysis of tests for paired lifetime data
1Department of Mathematics and Computer Science, University of Richmond, Richmond, VA 23173, USA. wowen@richmond.edu
Lifetime Data Analysis
|June 9, 2005
Summary
This study compares two tests for paired-data experiments using a new lifetime model. While one test is more powerful, the other may be preferred for practical reasons.
Area of Science:
- Statistics
- Biostatistics
- Survival Analysis
Background:
- Lifetime models are crucial for analyzing paired-data experiments.
- Exact parametric testing methods exist for comparing marginal survival distributions.
Purpose of the Study:
- To conduct a power analysis comparing two tests for a new lifetime model.
- To determine the more powerful test for paired-data experiments.
Main Methods:
- Utilized power analysis to evaluate two pivotal quantities for exact parametric testing.
- Calculated the power of each test under the new lifetime model.
Main Results:
- The power curves for the two tests were found to be very similar.
- The less powerful test may be advantageous due to pragmatic considerations.
Conclusions:
- The choice between tests involves balancing statistical power with practical utility.
- Further investigation into pragmatic factors influencing test selection is warranted.