Advances and Applications in Statistics
Volume 36, Issue 2, Pages 151 - 163
(October 2013)
|
|
ADJUSTING THE LASTER-JOHNSON-KOTLER NON-INFERIORITY TEST FOR COMPARING TWO PROPORTIONS
David Sotres-Ramos, Yoni Castillo-Tzec and Félix Almendra-Arao
|
Abstract: Non-inferiority tests for comparing two proportions are frequently used in active-controlled clinical trials. The method introduced by Laster et al. [9] offers a useful option for planning and interpreting active control trials when the selection of an absolute value (Blackwelder’s d) for clinical tolerance might be considered arbitrary orcontroversial.However, this Laster-Johnson-Kotler (LJK) approach, even when used with a continuity correction factor, has the disadvantage that its size is very different than the required nominal significance level (a). In this paper, we have adjusted this LJK procedure which yields a size very near to the required nominal level (a). |
Keywords and phrases: non-inferiority, dichotomous data, test size, Laster-Johnson-Kotler test, continuity correction. |
|
Number of Downloads: 403 | Number of Views: 1230 |
|