Related Experiment Videos
UMPU and alternative tests for association in 2 x 2 tables
1Department of Statistics and Operations Research, Sackler School of Exact Sciences, Tel Aviv University, Ramat Aviv, Israel. fuchs@post.tau.ac.il
Biometrics
|June 21, 2001
Summary
The uniformly most powerful unbiased (UMPU) test is not ideal for gametic association studies. Alternative adjusted-more extreme tables (A-MET) and equal-tails (ET) tests are proposed for better results.
Area of Science:
- Genetics
- Statistical Genetics
- Population Genetics
Background:
- The uniformly most powerful unbiased (UMPU) test was proposed as an alternative to Fisher's exact test for analyzing gametic association between polymorphic loci.
- However, the UMPU test is not suitable for two-sided alternatives, limiting its applicability.
Purpose of the Study:
- To evaluate the UMPU test for gametic association analysis.
- To address criticisms regarding randomized tests.
- To propose alternative statistical tests for improved power and applicability.
Main Methods:
- Comparison of the power of the UMPU test against other statistical tests.
- Discussion of criticisms related to randomized tests.
- Introduction of the adjusted-more extreme tables (A-MET) and equal-tails (ET) tests.
Main Results:
- The UMPU test is often less desirable in practical gametic association studies.
- The proposed A-MET and ET tests demonstrate better performance in certain scenarios.
- An extended p-value definition may mitigate some concerns about randomized tests.
Conclusions:
- The UMPU test is not universally appropriate for gametic association studies.
- The A-MET and ET tests offer viable alternatives with improved practical utility.
- Further research into extended p-value definitions could enhance the use of randomized tests.