作者: Paul Kabaila , Khageswor Giri
DOI:
关键词: Tolerance interval 、 Replicate 、 Frequentist inference 、 Confidence interval 、 Data set 、 Factorial 、 Statistics 、 Population 、 Mathematics 、 Factorial experiment
摘要: Consider a two-by-two factorial experiment with more than 1 replicate. Suppose that we have uncertain prior information the two-factor interaction is zero. We describe new simultaneous frequentist confidence intervals for 4 population cell means, coefficient 1-alpha, utilize this in following sense. These define cube expected volume (a) relatively small when zero and (b) has maximum value not too large. Also, these coincide standard obtained by Tukey's method, data strongly contradict illustrate application of to real set.