Example 8.4 To reduce the amount of shrinkage in an injection molded part, an experiment with four factors at 2 levels was performed. Only two interactions, A×C and A×D, were considered worthy of investigation. Therefore, the half fraction was chosen over the full factorial to conserve runs. A sketch of the molding scenario is shown in Figure 8.15. The experimental settings for each factor and the results of the 2^(4-1) fractional factorial design with 4 replications are shown in Tables 8.19.