Files in this item
|(no description provided)|
|Title:||Bahadur exact slope, Pitman efficiency, and local power for combining independent tests|
|Author(s):||Abu-Dayyeh, Walid Ahmad|
|Doctoral Committee Chair(s):||Marden, John I.|
|Department / Program:||Statistics|
|Degree Granting Institution:||University of Illinois at Urbana-Champaign|
|Abstract:||Many authors, for example, Fisher (1950), Pearson (1938), Birnbaum (1954), Good (1955), Littell and Folks (1971, 1973), Berk and Cohen (1979), and Koziol, Perlman, and Rasmussen (1988), have studied the problem of combining several independent tests. Many combination procedures have been proposed, and the relative performances of such procedures have been investigated using a variety of criteria, for example, power against specific alternatives, and Bahadur Exact Slope as the sample size of the individual tests increases, but the number of tests combined is constant.
In this thesis, we will look at the problem of combining p independent tests as p approaches infinity. We will look at a number of popular omnibus combination procedures, and we will compare their performances via Bahadur Exact Slope, Pitman efficiency, and local power in a number of situations. In these cases, it will be shown that no combination procedure is uniformly more powerful than the others, but the logistic method is generally more efficient than the others in terms of Bahadur Exact Slope, and the Inverse Normal method is better than the others when considering Pitman efficiency and local Power.
|Rights Information:||Copyright 1989 Abu-Dayyeh, Walid Ahmad|
|Date Available in IDEALS:||2011-05-07|
|Identifier in Online Catalog:||AAI8924752|