Robustness of Augmented Box-Behnken Designs to Two and Three Missing Observations
نویسندگان
1 Department of Mathematics and Statistics, Faculty of Physical Sciences, Alex Ekwueme Federal University Ndufu-Alike, Ebonyi State, Nigeria
2 Statistical Design of Investigations Unit, Department of Statistics, Faculty of Science, University of Ibadan, Nigeria
doi
10.22034/jirss.2024.1999151.1015چکیده
In real experimentation, second order response surface models may in some cases be inadequate and unrealistic due to lack of fit introduced by the presence of third-order terms in the response surface model. This willingly opens the door for augmentation of a second order response surface model to a third order model. The third--order design that takes care of the estimates of the model in this work is augmented Box Behnken Design (ABBD), which has always been confronted with missing observation scenario. This scenario that has affected the power of its test and desirable fundamental properties of this design, often surfaces when an observation is lost, ignored, miss-collected, etc. Therefore, this work constructed a minimaxloss design for ABBD that is robust to two and three missing observations under minimaxloss criterion.