The Use Of Taguchi Methods And Computer Techniques For Research And Process Improvement

The most common experimental approach of studying one variable at a time while holding all others constant is extremely inefficient. Use of statistical methods enables engineers and scientists to obtain meaningful information with the fewest experiments, reducing cost and wasted effort. In particular, the Taguchi method can be more efficient than traditional full factorial experiment designs, as it can study more variables with fewer nuts, and it is more robust against process "noise." Taguchi designs are also simple to formulate and the results can be analyzed by simple PC spreadsheets. The role and benefits of using the Taguchi method in experiments to develop and optimize processes is discussed.