ABSTRACT
Statistical experimental design (SED) is a method for constructing experiments that will mute the muddling effect of experimental error and increase experimental efficiency. SED leads to better and less-expensive data analysis. It is far superior to the classical one-factor-at-a-time experimentation often taught in school. Classical experimentation cannot account for variable interactions. Classical experimentation often leads to false conclusions, for example, that one has arrived at an optimal place, when in fact one has not. Figure 13.1 contrasts classical and SED methods. SED is a powerful tool used by too few engineers. This disuse is due to several factors. First, engineers and scientists can successfully (though not as efficiently) experiment without SED. SED is a power tool. In the hands of a skilled practitioner, it reduces the time for experimentation and squeezes the most information from the data.
