Related Experiment Videos
Development of an efficient SAS macro to perform permutation tests for two independent samples
G K Balasubramani1, Stephen R Wisniewski, Hongwei Zhang
1Epidemiology Data Center, Graduate School of Public Health, University of Pittsburgh, Pittsburgh, PA 15261, USA. balagk@edc.pitt.edu
Computer Methods and Programs in Biomedicine
|May 21, 2005
Summary
Permutation tests offer advantages over parametric tests for small sample sizes and unknown distributions, providing more accurate results. A new SAS macro simplifies these powerful statistical methods for clinical trial data analysis.
Area of Science:
- Biostatistics
- Clinical Trial Methodology
- Computational Statistics
Background:
- Parametric tests may be unreliable with unknown distributions and small sample sizes.
- Permutation tests offer improved Type I error control and accurate p-values in such scenarios.
- Computational complexity has limited the adoption of permutation tests in clinical trials.
Purpose of the Study:
- To develop an efficient SAS macro for performing permutation tests.
- To enable the use of permutation tests in clinical trial data analysis.
- To provide exact significance probabilities for various statistics.
Main Methods:
- Developed a SAS macro to generate all possible data permutations.
- Implemented permutation tests for comparing means of two independent samples.
- The macro calculates exact probabilities for a range of statistical measures.
Main Results:
- The SAS macro efficiently generates permutations and performs tests.
- Provides exact probability values for hypothesis testing.
- Supports analysis of diverse statistics including means, medians, and variances.
Conclusions:
- The developed SAS macro overcomes the computational barriers of permutation tests.
- Facilitates the application of robust statistical methods in clinical trials.
- Enhances the accuracy and reliability of clinical trial data analysis.