A goodness-of-fit test with nuisance parameters: numerical performance |
| |
| Institution: | 1. Charles University, Prague, Czech Republic;2. Technical University, Liberec, Czech Republic;3. Department of Biostatistics & Statistics, University of North Carolina, Chapel Hill, NC 27599-7400, USA;1. Merck & Co., Inc., Kenilworth, NJ, USA;2. Complete HEOR Solutions, North Wales, PA, USA;3. Lauschke Consulting, Morris Plains, NJ, USA;1. University of Pittsburgh School of Medicine, Pittsburgh, Pennsylvania;2. Pittsburgh Supercomputing Center, Carnegie Mellon University, Pittsburgh, Pennsylvania;3. Department of Mathematics, Soongsil University, Seoul, South Korea;4. Department of Mathematics, University of Tulsa, Tulsa, Oklahoma;1. HIV and AIDS Malignancy Branch, National Cancer Institute, Bethesda, MD, United States;2. CCR Collaborative Bioinformatics Resource, Center for Cancer Research, National Cancer Institute, National Institutes of Health, Bethesda, MD, United States;3. Advanced Biomedical Computational Sciences, Frederick National Laboratory for Cancer Research, Leidos Biomedical Research, Inc., Frederick, MD, United States;1. Department of Experimental Physics, Ural Federal University, Ekaterinburg, Russia;2. Institute of Metal Physics UB RAS, Ekaterinburg, Russia |
| |
| Abstract: | Jurečková and Sen (J. Statist. Plann. Inference 91 (2000) 377–397) proposed goodness-of-fit tests for models admitting nuisance location or nuisance location and scale parameters, based on the difference of two estimators of the location parameter, that are asymptotically first-order equivalent iff the null hypothesis is true. We illustrate here the numerical performance of these tests on simulated data, demonstrating their applicability to practical problems. Comparisons are also made with the well-known Shapiro–Wilk goodness-of-fit test. |
| |
| Keywords: | |
| 本文献已被 ScienceDirect 等数据库收录！ |
|
