Design and statistical analysis of confounded factorial experiments

Abstract

<p>Ten confounded factorial experiments were designed with the purpose of filling in the existing gaps among the available plans. The method of analyzing such experiments was presented and illustrated by the 5 x 2 x 2 experiment. For each design, the reasons why a particular plan was adopted were discussed; the instructions for computing the sums of squares and for constructing the summary tables were given; and the extent of confounding was determined.</p>