Summary: | 碩士 === 東海大學 === 統計學系 === 103 === In a screening experiment, the identification of active factors may go wrong due to the complex aliasing structure. With the use of means models, the MBMM(modified Box-Meyer method)was proposed to overcome this problem. In this work, we use the MBMM to the analysis of screening experiments with binary data. When the data are binary, marginal distribution cannot be obtained in a closed form. We use the Monte Carlo integral to compute the marginal distribution. Finally, a real example, the tile experiment, is analyzed by using our method.
|