Download the Article (PDF, 66 KB)

**Doug Zahn**

Professor, Florida State University, department of statistics

I teach STA 5126: Applied Statis-tics, an algebra level introduction to applied statistics for nonstatistics graduate students. I have been teaching and developing this course for the past five years, gradually molding it into one built around a Six Sigma term project. The text I use is *Statistical Thinking: Improving Business Performance* by Roger Hoerl and Ron Snee (Duxbury, 2002).

I have been thinking about how introductory statistics courses can be more effective since I began teaching statistics courses at Florida State in 1969 and came up with the following necessary ingredients:

- The students must see the potential value of the course on the first day of class.
- The course must be centered around a student chosen activity that he or she wants to improve.
- The course must engage students and steadily reveal more value.

The course structure I now employ allows me to incorporate these ingredients to produce results beyond what my students and I had been able to produce previously.

I begin the course with a “wanted” conversation to identify what the students want to get out of the course.(1) I explore their initial statements of wants, which are often broad and vague, such as, “I want to learn why statistics is required,” or “I want to learn more about statistics.” Together, we develop specific conditions for satisfaction for the course, such as, “By the end of the course I want to have completed one Six Sigma study on some aspect of my biochemistry laboratory procedures and to have improved that aspect.”

**Project assignment one: goals.** The students each list two contributions they want to make to their careers, how they think the course can help them do this and the results they want to get from the course. The remaining wanted conversations focus on their initial lists of desired course results. My goal is to convert these desired results into specific, measurable conditions for satisfaction.

The students use chapters one and two in *Statistical Thinking* to become more aware of what statistical thinking has to offer and to expand the possibilities they see for the course.

The statistical thinking process is the best way I have found to consistently produce results like this. Over the years, one of the most challenging questions for me has been, “Why is this course required?” I respond by:

- Asking what aspect of life the student is interested in improving.
- Identifying the processes involved and the variation in those processes.
- Exploring the value of understanding the variation and reducing unwanted variation.
- Promising the course teaches a strategy for doing all this.

**Project assignment two: define.** The students use chapter three to develop the language and concepts necessary to discuss processes. They each pick a process they want to improve and gather five baseline observations on the main output variable they hope to improve and on two input or process variables they think may relate to variation in the output variable. I coach them carefully on this assignment to clarify both what they are studying and what they hope to get out of the project.

**Project assignment three: measure.** The students operationally define their variables and develop measurement processes for each, indicating what steps they are taking to produce high quality data. Then they gather 10 more observations on these variables.

**Project assignment four: analyze.** The students read, discuss and explore the tools and concepts in chapters four, five and six to develop initial data analysis skills. These skills are applied to their 10 or 15 initial observations.

**Project assignment five: improve.** The students use chapter seven to learn how to design and analyze experiments. They each design a two-by-two factorial experiment with two replications to further investigate their process. They then take two more observations at the most promising factor combination to see if the results of the experiment can be replicated.

**Project assignment six: control.** The students study the tools and concepts in chapters eight and nine to learn how to perform and understand hypothesis tests and confidence intervals. They gather 10 more

observations under the promising conditions discovered in the improve step and use these inferential tools to compare the 10 observations to the 15 gathered in the define and measure steps.

**Project assignment seven: oral report.** The students present 10-minute summaries of their Six Sigma projects.Most student projects relate to individual improvement projects. Some are personal: increase running distance, lose weight, reduce body fat percentage, gain strength or reduce caffeine intake. Some are professional: improve reading process, improve key-wording process, study more effectively or improve transition process between classes. Though most students choose to do a project focused on improving some aspect of their lives, one or two each semester choose to study how to better pop popcorn.

Approximately 75% of these students have taken an undergraduate statistics course they profess to remember little of and from which they gained virtually no appreciation of how to use statistics in an interesting way. Virtually all the students completing my course are delighted they can now use statistics to improve anything they really care about.

As one student puts it, “Before this class, I took a statistics class offered by my department. At that time I did not like statistics a lot, so I decided to study qualitative research for my dissertation. After this class, my mind has been changed. I think I can also use quantitative research.”

In my experience, the effectiveness of these innovations is comparable to penicillin’s performance in Alexander Fleming’s first petri dishes: It works! Admittedly, my work is a pilot study, with all the expected questions: What is the impact of the instructor on this process? What are the necessary and sufficient characteristics required by an instructor to make this process work? What will it take to ramp up this course to other instructors in other contexts?

Of course, there are also barriers to the diffusion of these innovations in higher education:

- The academic reward system, especially at four-year colleges and universities, is still heavily skewed toward traditional research and grant activity. The faculty teaching statistics courses at these institutions are generally unaware of what process thinking is and thus don’t value it and can’t see the use of discussing it in an introductory course.
- Many faculty members don’t think the contribution of statistics is to improve processes via statistical thinking and methods. Many still have no experience in using these tools on any problem of personal interest to them. Thus, they continue to think a cross-sectional approach to teaching introductory statistics is the only way to go.
- Faculty and students expect classes to be run by a teacher in command mode who just delivers lectures.(2) Messing with these paradigms creates trouble on both sides of the desk.

We have work to do if we hope to increase the diffusion of these innovations. We must help our colleagues and students become change masters(3) and give them the opportunity to develop process thinking skills.

Note: My STA 5126 term project instructions and course syllabus can be found at http://stat.fsu.edu/~zahn/.

**References**

- D.R. Boroto, D.A. Zahn and D.C. Short, “A Theory of Consultancy,”
*Academy of Human Resource Development Conference Proceedings*, 2000, pp.1077-1084. - Daniel Goleman, Annie McKee and Richard Boyatzis,
*Primal Leadership: Realizing the Power of Emotional Intelligence*, Harvard Business School Press, 2002. - Rosabeth Moss Kanter,
*The Change Masters*, Simon & Schuster, 1983.

**Gregory H. Watson**

Managing partner, Business Systems Solutions Inc.

While there are examples of Six Sigma programs in universities, the impact of Six Sigma on the university has lagged behind its impact on business. This is not surprising because university educational curricula typically delay the development of ideas in society. Why?

Universities focus on education, and businesses concentrate on training. Education combines theory and potential practice, and training focuses on application and practice. The importance of practical applications is more abstract in the field of education where theory rules over practice, while in business the application is the driver in training. It’s like the proverbial professor’s comment, “I don’t care if it works in practice, does it work in theory?”

This observation is particularly valid when it comes to Six Sigma. Many of the learning elements of Six Sigma have been included in university education programs, including basic statistics, problem solving, design of experiments (DOE) and statistical process control (SPC). This, however, does not guarantee the way these tools are deployed in Six Sigma is served by the normal structure of a university curriculum.

The Six Sigma approach requires a greater integration of these tools than is found in university courses, and several aspects of Six Sigma are typically not included in an engineering or business degree program, such as defining operational problems, determining cost of operations and conducting project reviews.

How should a university program be structured to deliver Six Sigma education? Several constraints to educational programs inhibit using the same program as applied in industry:

- A three- to five-hour course does not provide enough contact hours to teach both theory and practice of Six Sigma.
- The sequence of courses must follow the sequence of learning applied in the Six Sigma define, measure, analyze, improve and control (DMAIC) model. University students are typically allowed to elect which courses to take to meet their needs.

The Six Sigma program, on the other hand, follows a rigorous process and application of tools that has been developed over a long time and is based on sound business practices. - The program of instruction relies on a series of educational specialists to instruct the individual courses with responsibility placed on the student to integrate the lessons across the classes into a holistic perspective of the discipline.

Business, however, uses a common approach to solving problems, and it becomes a shared language among its participants. This has been found to be superior to having many unique, albeit creative, approaches. Six Sigma standardizes an organization’s approach to problem solving by ensuring the logical step-by-step analytical process is consistently applied across problems. This requires a high degree of integration and teaching consistency across the courses in a Six Sigma program. Most universities will find this difficult to accomplish.

Shon’s Paradox in learning a new proficiency says students cannot initially understand what they need to learn; they can only learn by educating themselves and beginning to do what they do not yet understand.1 This is the problem with Six Sigma: Profound knowledge of the integration of the analysis tools and problem solving methodology is obtained through real-world practice.

So how could a program of study be structured for successful delivery at the university level, despite these difficulties? Here is my recommendation for the BB certification body of knowledge (see www.asq.org for a complete description). This program has been designed with the following guidelines in mind:

- Course objectives are based on real-world objectives, and the content is aligned to emphasize Six Sigma thinking rather than theory or numerical calculation.
- The program is designed for experiential learning with less emphasis on lectures.
- The program begins by teaching how to make a scientific inquiry using the DMAIC method followed by teaching applicable individual statistical tools.

The following topics should be divided to fit either a two-semester or three-quarter course sequence as a program of instruction. The number following the topic indicates the number of 50-minute instructional units dedicated to the topic, for a total of 90 units.

- Overview of Six Sigma.
- Six Sigma thinking (2).
- Lean thinking.
- Case studies in Six Sigma implementation.
- Managing with Six Sigma.
- DMAIC problem solving.
- Define, measure, analyze, design, validate (DMADV) product and process design.
- Six Sigma project management.
- Six Sigma statistical measures (2).
- Six Sigma problem definition.
- Introduction to statistical computing with software like that from Minitab.
- Descriptive statistics (2).
- Customer requirements analysis.
- Process analysis (2).
- Failure analysis.
- Measurement system definition.
- Measurement systems analysis (2).
- Process capability studies (2).
- Benchmarking.
- Measure case study (8).
- Multi-vari studies.
- Hypothesis testing (2).
- Analysis of variance (3).
- Regression analysis (2).
- Analyze case study (6).
- Experimental planning (2).
- Experimental design: factorial (3).
- Experimental design: fractional factorials (3).
- Experimental design: nested designs (2).
- Experimental design: blocking.
- Taguchi designs (2).
- Experimental design: curvature (2).
- Sequential experiments (2).
- Evolutionary operations.
- Improve case study (6).
- Control plans.
- SPC (2).
- Mistake proofing.
- Work simplification.
- Total productive maintenance.
- Quality system management.
- Statistical tolerance analysis.
- Control case study (6).
- Financial benefit realization.
- Project closure case study (4).

This program requires a detailed case study that allows students to track a Six Sigma project from beginning to end. While this approach to teaching Six Sigma falls short of the project based approach, it does closely follow the basic DMAIC methodology. To teach this program, the instructor must have real-world experience in these tools applied in the DMAIC context, not in an individual standalone application.

**Reference**

- Neil R. Ullman, “Statistical or Quantitative Thinking as a Fundamental Intelligence,” presentation at the Joint Statistical Meeting, Orlando, FL, 1995.

**Joseph G. Voelkel**

Associate professor and department chair, John D. Hromi Center for Quality and Applied Statistics, college of engineering, Rochester Institute of Technology

I decided to restrict the issues involved by discussing the influence the Six Sigma movement is and should be having on teaching Six Sigma at universities.

I also restricted my focus to universities in Michigan because of the clear role such universities play in the auto industry. Also, there is no question many universities in Michigan have responded vigorously and effectively to working directly with corporations in delivering noncredit Six Sigma training, but my interest was in seeing how much effect Six Sigma has had on credit bearing classes.

After selecting those universities that are either large public universities or ones I thought would have strong affiliations with the automotive industry, I ended up with a list of 10: Eastern Michigan University, Ferris State University, Michigan State University, Michigan Technological University, Northern Michigan Univer-sity, Oakland University, University of Michigan, University of Michigan-Dearborn, Wayne State Univer-sity and Western Michigan University.

I then entered each university’s website and searched for the phrase “Six Sigma.” Note: A search for “Six Sigma” entered in quotes will find phrases “six sigma,” “Six sigma” and “six-sigma,” but not “Sigma Xi has six members” or “6 sigma.” Also, the search was done using the search engine within the site, not a search engine directed to the site. For example, a search of the University of Michigan’s site, www.umich.edu, will search sites such as www.bus.umich.edu, the URL for the business school at that university.

I was able to find only three courses among these 10 universities that had “six sigma” in the title or course description:

- One business school offers a course called Six Sigma: Quality Tools for Operational Excellence. The course description says it focuses on SPC, DOE and acceptance sampling.
- One industrial engineering department offers a course in quality engineering that has Six Sigma in the course description, “ … the Six Sigma problem solving methodology.”
- One quality department offers a course called Fundamentals of Six Sigma, and the course description says, “Six Sigma coursework will focus on the application and use of practical problem solving tools and data interpretation to reduce costs through waste elimination. The course provides more than adequate preparation for the ASQ Black Belt certification exam.”

Of these three courses, only the third course seems dedicated to Six Sigma. If these 10 fine universities have not generally incorporated Six Sigma into their courses, we should not anticipate many other universities have done so either.

So what influence should the Six Sigma movement have on teaching at universities? It is important to distinguish the differing roles of the university and the corporation in teaching. In a university, the role of teaching is primarily education, which stresses the embodiment of knowledge and critical thinking. In a corporation, the role of teaching is primarily training, which is designed for practical application and for increasing a person’s skill set.

In a for-credit course at a university, a student might be expected to compare different approaches—Six Sigma, lean and total quality management—and stimulate and sharpen critical thinking to understand when and where the various approaches are most advantageous.

A training program at a corporation, on the other hand, might focus solely on Six Sigma as the approach that will produce the best results and train participants in the Six Sigma process and tools. Given this distinction between education and training, I am not discouraged by the lack of formal Six Sigma training at universities.

Let’s look at it in more a specific way. First, in my ideal world, some universities would provide undergraduate programs that prepare students to do well in both a career and life. For a career that includes process improvement, universities would start by offering students courses in basic statistics that teach both statistical thinking (see www.asqstatdiv.org/stats-everywhere.htm) and statistical methods.

Additional courses would be required to expand the knowledge base for the student’s chosen career field. For engineers this would include at least measurement systems analysis, SPC and some DOE in addition to the base engineering curriculum. One or two courses would not make students knowledgeable enough to perform sophisticated work but would show them what is possible. The knowledge acquired from these courses would be valuable to students even if they don’t become employed by a Six Sigma corporation.

Also in my ideal world, corporations should train their employees in process improvement systems such as Six Sigma. The result is more uniform knowledge, a better understanding of corporate culture and commitment and the camaraderie and cross functional communication that accompanies such training.

Finally, employees who want more knowledge after this training could return to a university to take classes or earn a certificate or master’s degree. They would benefit from having the background and motivation that has been provided by working at the corporation. Some universities would even offer a graduate certificate program in Six Sigma that includes the ASQ Six Sigma body of knowledge and a focused process improvement project, with an emphasis on education rather than training.

Corporations would continue to sponsor some of their employees who want to attend for-credit programs because it enhances the training programs. Universities would offer more advanced, but applied, programs, such as a master’s degree in applied statistics.

I know such a strategy can work based on our graduate level applied statistics programs at Rochester Institute of Technology. Many of our graduate students have provided leadership in the development of Six Sigma programs in their corporations. And those who have earned BBs or Master BBs (MBBs) come to us thirsting for a deeper understanding of topics such as DOE, reliability and robust design, all of which we provide in online courses. Students also benefit from contact with instructors, who are far more knowledgeable in the area than most corporate trainers. I have little doubt other schools with such programs have seen similar success.

**Angie Patterson**

Statistician, GE Global Research, and adjunct professor, department of statistics, Virginia Tech

Over the past year, a number of schools have begun to recognize the value formal training in Six Sigma will bring to its students and are taking action by adding a Six Sigma course to their curriculum. I am delighted to see this happening and recognize it as the first wave of the Six Sigma movement within universities.

To my knowledge, universities have not yet documented the fact benefits gained by applying Six Sigma to improve the quality and productivity of processes are not limited to the corporate world. After all, budget cuts and downsizing are just as much an issue in universities as they are in corporations. Isn’t this an opportune time to leverage Six Sigma to bring more efficiency to university processes? This internalization of Six Sigma is what I refer to as the next wave of the Six Sigma movement within universities.

To my knowledge, Virginia Tech was one of the pioneers of the first wave, offering a graduate level course in Six Sigma through the department of statistics in the fall of 2002. The motivation for this course came after the department realized graduate students need to begin building the skills necessary to become statistical leaders of the future because expertise in statistics is no longer sufficient for long-term success as a professional statistician.

For students of statistics, this requires augmenting their knowledge of statistical methods with skills, such as statistical thinking through structured problem solving, team facilitation, the ability to relate statistical results to bottom-line goals, project planning and management and an introduction to quality improvement tools. Another critical skill is the ability to effectively communicate with nonstatisticians.

The evolution of Six Sigma in corporate America has underscored the fact it doesn’t take a degree in statistics to become a statistical leader. Instead, many MBBs and BBs have degrees in other fields, including engineering, science and business. For this reason, a university course in Six Sigma is equally beneficial to students from departments outside statistics because it will help prepare them for leadership roles. Schools such as Arizona State University offer formal training in Six Sigma that is open to students of engineering, business and statistics.

The course at Virginia Tech encourages enrollment from students outside statistics to facilitate collaboration between statisticians and nonstatisticians. It is based on the recommended curriculum in Roger Hoerl’s article, “Six Sigma Black Belts: What Do They Need to Know?” published in the *Journal of Quality Technology* (Vol. 33, No. 4). Students get hands-on experience applying the Six Sigma roadmap and tools and learn team facilitation and leadership skills by completing a team project over the course of the semester.

The next wave of the Six Sigma movement in universities can’t happen soon enough. With budget cuts and downsizing of faculty and staff, universities are already being asked to do more with less. What better time to improve the efficiency and productivity of university processes?

The challenges facing the next wave are two-fold:

- Lack of Six Sigma examples to which universities can relate and be motivated by.
- Lack of faculty resources to lead Six Sigma projects.

We are beginning to address these challenges at Virginia Tech. Last fall, the university provided students with a real project through a partnership with General Electric and a simulated business project. This fall, in addition to the simulated project, the students are undertaking a real project within the university, the goal of which is to improve the efficiency of the graduate admissions process.

Besides meeting the goals of the graduate school, we hope to generate a success story that can be shared with and translated to other areas in the university. We also plan to share this project with other universities.