Abstract
This article gives the way to assign six-level factors to the L36 orthogonal array. Firstly, it is found that three six-level factors which are orthogonal to each other can be assigned. In the case of two six-level factors, two three-level factors can be also assigned additionally. Four three-level factors can be assigned with one six-level factor. Secondly, if we admit the least confounding between the assigned main effects, we can assign at most six six-level factors to L_<36>. At that time, all main effects are estimable by using regression model with dummy variables. The optimal design is determined based on D-optimal criterion. The way of assigning several six-level factors is shown by nearly orthogonal arrays. The remaining two-level columns are available for assignment of two-level factors in each case.
