| Summary: | 碩士 === 國立高雄師範大學 === 數學系 === 94 === In every kind of experiment, we hope that the experiment may offer more information under taking little money and time. In experiment design, we expect some main effects and two-factor interactions could clear to estimate after adding quarter replicate of the origin design in the combined design. The double semifolding design（denoted fo=A, ss=B, le=C）may perform this kind of experiment . We use 2^(k-p) fractional factorial designs and apply Excel VBA to try many cases to find the results. At the beginning, we fold one factor at the level of the other factor and then we search its corresponding subset to have the semifolding design. Forthermore, we expect that more clear effects can be found in the combined design. Finally, we try to find the rules for our searches.
Because only quarter replicate add to the original design, we can not obtain more many clear main effects or two-factor interaction in the combined design. We hope that some results in this article can help the experimenter for particle use.
|
|---|
