W. I. Notz, professor of statistics, Ohio State University, and editor of Technometrics
Statistics is a significant component of the Six Sigma initiative. Black Belt (BB) training includes many statistical techniquesbasic inference, regression, analysis of variance, design of experiments, measurement and process capability, and statistical process controlbecause these are useful tools for solving problems. However, many practitioners of Six Sigma were not initially trained as statisticians; the Six Sigma initiative was developed primarily by engineers and managers. This raises a number of questions:
As a professional statistician, I may be biased, but I believe statisticians have much to offer practitioners of Six Sigma. This includes:
1. Up-to-date research. Practitioners seek solutions to specific problems. Researchers explore the best ways to solve problems. As a follow-up to any real project you face, consider investigating what the latest research says about how you might solve a similar problem in the future. ASQ journals as well as many intended for statisticians (for example, The American Statistician) include both expository and technical articles on new and improved statistical methods.
I would recommend scanning the titles and abstracts in each issue. If an article appears to address a problem that you face in practice, read the introduction and any examples.
2. Software for implementing the latest methods. Such software is often mentioned in research articles and may be available from the author. Unfortunately, this software is not always in a user-friendly form, so be prepared to invest time in learning how to use it.
3. Informal advice. If you have a good relationship with a statistician in your company, a former teacher or a statistician you met at professional meetings, this person may provide informal advice about unusual problems.
4. Training and resources for more advanced statistical techniques. W.H. Woodall, editor of Journal of Quality Technology, makes this point very clearly, including several recommendations, in the February issue of Six Sigma Forum Magazine.
Practitioners of Six Sigma also have much to offer statisticians, including:
1. Special nonstatistical skills. BB training includes problem solving skills beyond those taught in statistics courses. Statisticians do not necessarily have formal training in communication skills, managing projects, developing teams or assessing progress. Statisticians can benefit from partnerships with Six Sigma practitioners in the projects they face.
2. Improvements in the training of applied statisticians. Some of the nonstatistical material that is part of BB training might be incorporated into the training of applied statisticians in academia. If these skills make statisticians more effective and more attractive to industry, then academic programs are doing their students a service by providing training in these skills.
3. Sources of research problems. Some of the best research is motivated by real problems. However, academic statisticians often have little or no experience with real problems. By sharing case studies, Six Sigma practitioners provide re-searchers with problems that can serve as motivation for new and more effective statistical methods.
4. Case studies for teaching. The teaching of statistics at all levels is enhanced by good case studies. Descriptions of interesting projects will be useful to statistics teachers.
This is by no means an exhaustive list of ways in which statisticians and Six Sigma practitioners can benefit from one another. My point is to encourage cooperation, not competition, between statistics and the Six Sigma initiative. I welcome your feedback.