| Summary: | 碩士 === 國立高雄師範大學 === 數學系 === 91 === After finishing an experiment, based on the information the experimenter has gotten, he might decide to perform a follow-up
experiment to gain more clear effects about important factors.
Although adding a foldover design of the same size as an original experiment by reversing the signs of one or more factors is commonly used for follow-up experiment, under some specific conditions a half-fraction of the foldover design named semifolding will suffice. These specific conditions were given and proven for semifolding 2^{k-p} fractional factorial designs. On the other hand, if the experimenter has some prior information about specific factors and wishes to obtain more relevant clear effects, he can use the semifolding design to isolate important effects in the 2^{k-p} fractional factorial designs. This article also discusses the optimal semifolding designs for one's need by means of exhausting all possible semifolding designs for a given design. Moreover, the semifolding designs of selected 8-run and 16-run fractional factorial designs are constructed and tabulated for practical use.
|
|---|
