dc.contributor.advisor | McCann, Melinda H. | |
dc.contributor.author | Grant, Yanina | |
dc.date.accessioned | 2014-04-15T22:26:39Z | |
dc.date.available | 2014-04-15T22:26:39Z | |
dc.date.issued | 2006-05-01 | |
dc.identifier.uri | https://hdl.handle.net/11244/9493 | |
dc.description.abstract | The objective of this project is to derive two simultaneous inference approaches, Dunnett and Jeffreys-Perks, to quantify proportion differences between a control and several non-control levels under a group testing scenario. Both procedures were evaluated in terms of estimated coverage probability via simulations. Ten thousand simulations were run for diverse data scenarios and the coverage probability was estimated by calculating the proportion of times the confidence intervals simultaneously contained the real proportion difference. Findings and Conclusions: It was found that when using large group sizes of about 15 or more observations both the Dunnett and Jeffreys-Perks approaches perform fairly well. For small sample sizes, between 5 and 10 observations, the Jeffreys-Perks intervals are recommended as their coverage is closer to the nominal 1-α level, and it also is a conservative approach while the Dunnett procedure is anticonservative. | |
dc.format | application/pdf | |
dc.language | en_US | |
dc.publisher | Oklahoma State University | |
dc.rights | Copyright is held by the author who has granted the Oklahoma State University Library the non-exclusive right to share this material in its institutional repository. Contact Digital Library Services at lib-dls@okstate.edu or 405-744-9161 for the permission policy on the use, reproduction or distribution of this material. | |
dc.title | Multiple Comparisons with a Control Under a Group Testing Scenario | |
dc.type | text | |
dc.contributor.committeeMember | Monks, Stephanie A. | |
dc.contributor.committeeMember | Warde, William D. | |
osu.filename | Grant_okstate_0664M_1729.pdf | |
osu.college | Arts and Sciences | |
osu.accesstype | Open Access | |
dc.description.department | Department of Statistics | |
dc.type.genre | Thesis | |