The Effect of the Joint Violation of Two Assumptions on the Size of the F-Test Under Permutation for the Randomized Block Design

作     者：Toothaker, Larry E. 

作者机构：The University of Oklahoma Norman Oklahoma 73069 455 West Lindsey Street Room 705 United States 

出 版 物：《Journal of Statistical Computation and Simulation》 (J. Stat. Comput. Simul.)

年 卷 期：1972年第1卷第1期

页      面：55-64页

学科分类：0202[经济学-应用经济学] 02[经济学] 020208[经济学-统计学] 07[理学] 0714[理学-统计学（可授理学、经济学学位）] 0701[理学-数学] 0812[工学-计算机科学与技术（可授工学、理学学位）] 

摘      要：The present study attempts to explain the problem encountered by Baker and Collier (1968) of a “humping of empirical size of Ft under permutation for the simple randomized block design for combined levels of variance heterogeneity and block-treatment interaction. Values of corrected nominal size were calculated with the degrees of freedom corrected by two methods: one to correct for variance heterogeneity alone and another to correct for the change in variance heterogeneity induced by the block-treatment interaction. The second method yielded nominal values of size which showed the best agreement with empirical values of size having the “humping but did not exactly duplicate the peculiar pattern. © 1972, Taylor & Francis Group, LLC. All rights reserved.