To provide sound answers to specific problems, it is possible to design appropriate test programs using methods developed by mathematical statisticians. A simple factorial design, as applied to the study of green compressive strength of a foundry sand mix, is outlined and discussed in some detail. The efficiency with which conclusions are drawn from the test results is attributed to the fact that factorial and fractional factorial designs enable the test engineer to use every observation in making each conclusion. Furthermore, the same designs provide an opportunity for examining the behavior of one variable at more than one level of each of the other variables included in the test program. Some simple methods of numerical analysis are applied to the test data preliminary to their interpretation.