In unreplicated 2 k-p designs, the assumption of constant variance is commonly made. When the variance of the response differs between the two levels of a column in the effect matrix, that column ...
Results that may be inaccessible to you are currently showing.
Hide inaccessible results