Design of experiment at use of logit model of missingness

The problem of optimum design of experiment for multiresponse regressions models in a situation when outcomes of experiments contain missing values is considered. The probability approach to exposition of missingness mechanism is used; Association of Probabilifies of structures Of missingness on input variables is modelled with use of a logit regression. Design of experiment for optimization of properties of parameters both the basic regression, and logit model is considered.