298 – Experimental Design and Statistical Engineering: Estimation with Constraints
Yates' Method for Computing Sums of Squares for Nonreplicated Experiments
Mary Marion
Yates Method is used to compute the effects of k factors each at two levels on a response variable. This calculation includes the sums of squares and coefficients in a factorial model for nonreplicated experiments. An overview of the Yates algorithm is presented along with different types and uses of Yates inputs and expected outputs. The estimated effect estimates are ranked and used to identify potential models which are validated by analyzing the residuals. Graphical techniques are illustrated and the statistical performance of different approaches is compared. Various formulas for variances are developed using concepts from mathematical statistics.