Ethical Behavior Not Shared By All
I read March Laree Jacques' article on ethics (September 1999, "The Call of Quality: Doing Right Things Right," p. 48) with personal interest. In August, I terminated my position as director of quality assurance with a large aerospace manufacturer of pyrotechnic devices because I had reason to believe the company's business practices were unethical. The manufacturer was being investigated by the FBI, the Department of Defense and the Environmental Protection Agency.
When I applied for unemployment, it was denied. My hearing in October did not change the state's opinion. I even referenced Jacques' article during my hearing as an example of the standards ASQ sets. It appears that my state--at least its unemployment office--does not share ASQ's belief in ethics. Its position was that I should have continued to function as the head of quality even though I knew records were being falsified.
It appears that the bottom line here is that my state values $230 a week more than an honest quality professional.
STEPHEN P. DEMSKI
Introduction to Variability And Replication
I enjoyed reading the article "Burn the Brownies" by Lynne B. Hare (August 1999, p. 92), and I agree with him on the importance of experimental design in industry and science. However, I take issue with some aspects of the experiment Hare conducted with his daughter, Sarah. He missed a golden opportunity to introduce Sarah and his readers to the notion of variability and replication.
In the case of the experiment performed by Hare, variability is associated with differences among the eight boxes of brownie mix and the inability to repeat the factor settings exactly from one run to the next. It would have been better to buy 16 boxes and have each factor setting replicated twice. Then Sarah, with the help of her father, could estimate the experimental error variance rather than rely on the assumption of no three factor interaction.
I know that replications are not possible in many industrial experiments, but replications would have enhanced the validity of the investigation in this small experiment.
Thanks to Klaus Hinkelmann for his insights. I agree with his comments, but in Sarah's case--as in the case of most people just beginning to learn the effective techniques of experimentation--it is important to keep the efforts as simple as possible. Sixteen experimental combinations would have destroyed the kitchen even more than the eight we originally ran. Sarah's patience would have been exhausted as well.
As a result of this experiment, Sarah's understanding of the method and, consequently, her enthusiasm for its use grew. If we had had an opportunity to run a second experiment, we would have made it more elaborate.
This addresses a larger, more difficult situation for the consulting
statistician: Do we oversimplify experimentation for beginners
at the risk of missing important effects, or do we do the right
thing statistically and risk losing the client to the old practice
of one-at-a-time experimentation? During consulting sessions I try to get a sense of the client's understanding and tolerance and then design experiments accordingly. Other consulting statisticians may choose other paths; then we have statistician to statistician variation. Oh, no.
LYNNE B. HARE
East Hanover, NJ
Preventive, Corrective Case Study
It was great to see Debra Dyer's letter ("Preventive, Corrective Definitions Do Differ," September 1999, p. 14) about preventive and corrective actions and the confusion associated with them. I hope the case study below is helpful to QP readers. I often use this case study for internal quality auditor and ISO awareness training. The case study is as follows:
In September 1994, the entire country of India was panicked because of a plague epidemic. The communicable disease was first discovered in the city of Surat. Surat is located in the western part of India and is known for its textile mills. During the period of one week, many people became infected, and hundreds of people died from the plague in Surat. The main causes of the plague were untidiness and rats.
To cope with the situation, the Surat administration admitted all of the people infected with the plague to various hospitals and nursing homes and banned the assembly and gathering of people. All the public meeting places were closed, and everyone in Surat was vaccinated.
The Surat Municipality brought forth a cleaning movement, and the citizens were educated on the importance of cleaning through electronic media, newspapers and posters. Concerned authorities reviewed all the garbage collecting places and took care to make the city clean and tidy.
Barring a few cases, the disease was limited to Surat. To avoid the spreading of disease to adjacent cities and other parts of India, the borders of Surat were sealed and migration was prohibited. The importance of cleanliness and tidy surroundings in preventing the plague was stressed throughout India. Local bodies in all the districts made every effort to make the place clean. With all these efforts, the plague was controlled in two weeks.
After trainees study the case, I ask them to reply to the following questions:
1. What was the nonconformity?
2. What was the root cause of nonconformity?
3. What was the disposition action?
4. What was the corrective action?
5. What was the preventive action?
Sultanate of Oman
Results Should Show What Was Actually Found
The purpose of this letter is to make some comments on the article "Two-Location Gauge Eval-uation" (Neal D. Morchower, April 1999, p. 79).
Unless I am missing something, the article says you might get a result that shows the location is significant, or you might get a result that shows the sequence is significant. The location and the sequence results are both shown in Figure 3, and the raw data is presented in Table 1. Thus, if you want to know the actual result, you must do the analysis of variance (ANOVA) yourself. This is kind of like reading a mystery novel and not being told the name of the murderer.
My main point is that I don't think it's good editorial policy/practice to have readers figure things out for themselves. Most of the working quality engineers I know are too busy to consider doing the analysis of variance themselves. Unless there is something specific that motivates the readers, some of the value of the article may be lost.
Figure 2 also shows the different results one might get--primarily via plots of averages or individual values. I wanted to know what was actually found, so I processed the data further. I ran ANOVA for operators and parts and parts and sequence number within the four location-sequence groups. One could show the plots as significance is found, but the author did not do that.
Regarding the primary analysis (Figure 3), the actual result numbers are very close to the location example. The location-sequence interaction was not significant. Because the results were so close to the printed ANOVA location example, I wonder if there is a slight change in the used data versus the published data?
In looking at parts and sequences within each of the four location-sequence groups, I discovered that sequence is only significant in the location 2, sequence 2 case. I concluded that without more support there is no strong suggestion of a sequence effect in the six runs. It is noted that the parts-sequence interaction was used for error because it's impossible to replicate the second trial.
The other actual result worth noting is that the operator-part interaction is significant at the 0.01 level in one group, at the 0.05 level for two other groups and at the 0.10 level in the fourth group. Operators are only significant in one group.
To recap, Figure 2 does a good job of indicating various result possibilities. I just think that the message would be much stronger if the results showed what was actually found.
It is apparent that a little more work is needed to pin down the cause of the location difference. One paragraph indicates that the test equipment was different at the two locations. It also states that swapping operators could be beneficial, but only if the test equipment was the same. It seems to me that the different test equipment is a good suspect for the location difference.
I commend Quality Progress for publishing an article of this type.
You got me! I'm sure you are aware that problems in life are rarely black and white, but often include several shades of gray. The actual results that inspired the article are no exception.
The answer to the mystery involves two additive main effects: a tester bias of about 2 rpm due to equipment differences and a run-in effect of about 1 rpm from trial one to trial two. For the sake of demonstration, I used a trick that my wife--a teacher--recommends: When teaching a complicated topic, start with easy examples. In this spirit, I provided examples of a location, a sequence and a run-in main effect. All examples reflect real situations I've encountered, but as you hinted, one might find any combination of these effects in main effect or interactive forms.
I tried to prove two points in the article, and I would like to reinforce them here. First, don't get hung up on the ANOVA and statistics. Instead, focus on graphical analysis as in Figure 1. Take this simple example: Motor A and motor B both initially operate at exactly 58 rpm. There is a run-in effect of +1 rpm for their second trials and a tester bias of +2 for the Anywhere tester. Plot the rpm of each motor with trial one on the X-axis and trial two on the Y-axis.
Test motor A first on Nowhere and then on Anywhere, while motor B has trial one on Anywhere and trial two on Nowhere per my technique in the article. In this scenario, motor A tests at 58 rpm and then at 61 rpm, but motor B tests at 60 rpm and then at 59 rpm! Imagine motors A and B as groups of motors with rpms ranging from 57 to 62 being tested in such a fashion. Plotting these results clearly demonstrates the power of graphical analysis and my population-splitting technique. The power of this technique is the second, but equally important, point I wanted to make in the article.
I am glad to see that you took the time not only to read the article, but also to analyze the data. I agree that few of our peers (myself included) would have the time or incentive to do so. Your efforts allowed you to discover the inconsistency between the actual results and the location example provided. For all your hard work I will reward you by pointing out another inconsistency: Your letter opens by telling Quality Progress that you "don't think it's good editorial policy ... to have readers figure things out for themselves," and your letter closes by "commend(ing) Quality Progress for publishing an article of this type." Perhaps this is an example of one of those shades of gray. Thank you for your interest, and I hope others got as much out of the article as you did.
NEAL D. MORCHOWER
Problem With Standards Is Communication
The July 1999 issue of Quality Progress included a very nice set of articles on standards, with perhaps the best being Susan Daniels' summary of standards ("Standards--Problem or Solution?" p. 27).
Over the last 25 years, I have been involved with technical or engineering specifications in a variety of ways including writing them and using them as a project engineer. I am currently working as a facilitator for a major specification update in our company. As Daniels suggests, you either love standards or you hate them.
Why the big debate? I think everyone would agree that standards or their subset specifications provide a great service. They are the first step toward maintaining the technical integrity of a facility. Without standards we have no chance of ensuring that we will end up with a quality product. So why the big fuss?
I think the issue isn't the content of these documents, but the lack of communication. Standards aren't fun to read--their very nature and layout make them inefficient communication vehicles. Over the last few months, numerous letters have been written complimenting Quality Progress on its new look and improved readability. The biggest obstacle to communicating through specifications is the specification itself.
The document we call a standard--even ISO 9000:2000--looks like it was created 50 years ago. With all the tools we have available to us today, we can create a new paradigm for standards in terms of design and layout. Tools of good graphic design are available across a much larger spectrum than ever before. Coupling them with hyperlinks, HTML documents and more, we can create a series of linked documents that are easy to navigate and eliminate communication barriers.
The purpose of a standard is to communicate. If we can see our standards in a new light, then we can improve the use and understanding of these documents. I believe this can result in a 1% savings, and a 1% savings of a very large number is still a very large number.
To design is much more than to assemble, to order or even to edit. To design is to add value and meaning, to simplify and to clarify. We have that capability and we need to use it.
JAMES R. HEMSATH
Eagle River, AK
Shift Toward Creativity Will Benefit Companies
An article in the June 1999 issue, "Quality Profession Must Learn To Heed Its Own Advice" by Thomas Pyzdek (p. 60), was well-written and contained a lot of historical references.
The struggle between control and creativity is my favorite business issue of the '90s. Some companies have prospered with intense control systems in this decade, with the marketing advantage of ISO 9000 pushing all of us into a bureaucratic maze. In the next decade most of these companies will disappear if they haven't shifted to creativity over control.
I manage manufacturing organizations, making them cheaper, better and faster for competitive advantage. I do this through developing and encouraging innovation in everyone in my organization. It appears contradictory when we say "improvement through age-old methodology." Sure enough, some innovative, new idea company always comes around and topples the corporate giant.
When I was forced to leave my job after 23 years due to downsizing, I learned the painful truth that "it can happen here." In the information age and with the rapid changes in technology, the 21st century won't be a safe place to stand still. Pyzdek was absolutely on target with his article--looking forward toward developing new, creative ideas on how to manage quality and business.
More, Better, Faster Authors Earn Slap On the Back
The authors of the article "More, Better, Faster From Total Quality Effort" (August 1999, David A. McCamey, Robert W. Boggs and Linda M. Bayuk, p. 43) should be given a big slap on the back for presenting an outstanding study on how cross functional teams and quality methods can lead to great results. Procter & Gamble was and is a leader in promoting total quality, and it's great to see an article on how to do it outside of manufacturing.
DONALD S. ERMER
Six Sigma Lacks Breakthrough Thinking
What is all the fuss about Six Sigma? In all the articles I read so far (October 1999, "Cowboy Quality," Miles Maguire, p. 27), I have yet to see anything that could be considered breakthrough thinking. On my desk, I have a brochure from a well-known consulting firm detailing the Six Sigma Master curriculum. I read it over twice and I still don't see anything more than the certified quality engineer (CQE) and certified quality manager bodies of knowledge.
What is it that justifies the current rage for Six Sigma? I reviewed some of the topics in the curriculum, and I don't think anything in there is new. For example, teamwork for success, attributes and variables, cost of quality, quality function deployment, process capability and failure mode effects analysis are some of the topics covered in the Master course. Are any of these ideas new? No.
I believe any CQE with proper management support is already trained to be a Master. I also believe that Six Sigma training will enrich many more consultants than quality programs.
PAUL A. IPOLITO, CQT,
CQA, CQE, Certified Quality Manager
Six Sigma Has Its Risks If Not Properly Applied
As a one-time student of Mikel Harry--I attended a General Electric Master Black Belt training session he taught--I was pleased to read the October 1999 article ("Cowboy Quality," Miles Maguire, p. 27) on his accomplishments and philosophy. As a veteran of Six Sigma implementation, I would like to emphasize some points that were made in the article and describe a few that were left out or minimized.
I want to re-emphasize the value and power of the total business approach to Six Sigma. I firmly believe that the vast majority of the benefits from Six Sigma come from establishing a framework of common language and methodology throughout the entire enterprise. As described in the article, delivering and driving this framework into the organization through the Black Belt program/project management structure and individual performance evaluations/bonuses are critical elements in the Six Sigma success story.
Yet, as with any attempt at cultural change, there is a risk associated with the forceful introduction of the Six Sigma methodology into business. I saw this risk manifest in two ways. In one scenario, the drive created by upper management to complete a targeted number of Six Sigma projects was so strong that many trivial activities were treated as formal Six Sigma projects. While there is merit in breaking large problems into smaller, more manageable pieces, the pieces in this situation were so small that the credibility of the program was damaged.
The second damaging situation I observed was one in which Six Sigma Black Belts were repeatedly singled out publicly as the only people who had a chance at a promotion. As a result, many people became Black Belts only to get their tickets punched. Meanwhile, those who did not or could not become Black Belts were left with feelings of resentment and disillusionment. Once again, the credibility of the program was damaged by people focusing on the metric rather than the mission.
The moral of this story is that the methodology and approach to Six Sigma as described by Mikel Harry are powerful and effective ways to create a high performance business culture. However, this very same power can cause more harm than good if not properly applied.
Cowboy Quality Just Won't Cut It
There's no mystique surrounding Six Sigma for those who understand total quality management (TQM). Deming and Shewhart no doubt rolled in their graves when they heard Mikel Harry say, "TQM was a great thing if you had 30 years to realize the gains from it." This is what Miles Maguire quoted Harry as saying in his article "Cowboy Quality" (October 1999, p. 27). I can only imagine what Juran, Feigenbaum and Crosby are saying--perhaps that big egos are as vulnerable as ever to flimflam artists selling the latest quick fix remedy. Sounds more like drugstore cowboy quality to me.
What I can't understand is that the American Society for Quality--the greatest advocate of quality control and TQM--was hoodwinked that easily. Six Sigma is a gimmick for those who don't understand or can't muster the courage to commit to the Baldrige principles or the principles of process management and continuous process improvement promoted by Deming, Juran, ASQ and others.
Statistical process control (SPC) involves monitoring the performance of a process to ensure that it remains within the upper and lower control limits, even out to six standard deviations. True six sigma is less than one part per billion, not two parts per billion. Can you imagine an SPC chart that requires less than one part per billion as the upper control limit? The average company operates at 6,200 parts per million, if that. Harry is right about one thing, though: Achieving this level of quality will take time. It took the Japanese a generation. There is no quick fix.
The people who buy into programs such as Six Sigma are people who can't look any further than the requirements of stock analysts. Stock analysts are not stakeholders. Baldrige suggests that senior leaders are responsible for "creating and balancing value for customers and other stakeholders." It implies that an organization needs to collect feedback from customers, employees, owners (stockholders, not stock analysts), suppliers and the public in order to identify the requirements so it can create and balance value for each group. However, the CEOs of the companies that embrace Six Sigma as a vehicle for quality improvement can't look beyond the brims of their cowboy hats to see the view of the future.
There's no shortcut to quality as far as I'm concerned. I understand quality as Juran, Deming, Shewhart, Feigenbaum, ASQ, the keepers of Baldrige and others teach it: It's just plain common sense. Simply identify stakeholders' requirements, develop processes to exceed those requirements and monitor the processes' performance using SPC charts. If the gap between the requirements and the ability of the process designed to meet them is dramatic, implement dramatic breakthrough approaches. Redesign the process or create a new process. Do whatever it takes.
Poor quality is expensive. True quality, quality that exceeds stakeholder requirements, is free. The return on it is about three times greater than the return on quality that does not meet stakeholder requirements.
I don't blame Harry for doing what he's doing. He is obviously laughing all the way to his big ranch in Arizona. However, I am disappointed in ASQ for endorsing such a thing.
'QP': Independent Or Promotional?
The October 1999 issue of Quality Progress raised some disturbing questions with the article "Cowboy Quality" (Maguire, p. 27) and the Six Sigma Academy/ASQ Alliance advertisement (p. 9).
First, one should ask whether a membership based organization such as ASQ should enter into an alliance with one provider of a given service over and above any other provider. The ad on page nine denigrates other providers of Six Sigma services, some of whom undoubtedly count ASQ members among their ranks. The alliance means that these members are forced to contribute to their direct competition!
A second potential consequence concerns ASQ members who engaged the help of an unendorsed provider only to have their judgement second-guessed by management aware of the ASQ endorsement. Isn't it crucial that ASQ maintain an even hand in such matters?
The feature article about Mikel Harry leads one to ask what kind of periodical Quality Progress aims to be. Is it supposed to be an independent journal for the quality profession, or is it supposed to be mere promotional literature on behalf of a business?
While Harry deserves considerable credit for raising awareness and promoting the Six Sigma philosophy and methodology, the article appears to do little more than promote the academy. Mikel Harry has yet to demonstrate the long-term contributions to society that would warrant Maguire's comparing him to such gurus as Walter A. Shewhart and W. Edwards Deming.
If Quality Progress wishes to maintain the appearance of an independent journal, any articles written concerning a venture in which ASQ has a financial interest should at least include a disclosure of the terms of that interest. That way readers can judge the degree to which a conflict of interest may color the reporting.
ROBERT H. MITCHELL, chair
JOHN W. MORAN, chair
Quality Management Division
This letter was sent on behalf of the Statistics Division and the Quality Management Division councils and was signed by the divisions' officers.
For the last several years, the American Society for Quality's strategic plan has contained a provision calling for ASQ to pursue collaboration with external partners. This strategy was developed to meet the Society's objective of being "a worldwide provider of information and learning related opportunities related to quality." The strategy also serves to make the society more agile in responding to the needs of its members and customers.
Such partnerships have become increasingly common in today's business environment, and we believe they are vitally important because they can greatly extend the range of offerings that the Society can provide on its own while allowing members to gain access to information, training and quality tools at lower costs. They represent a win-win situation (see "Partnering Pays Off," May 1999, p.6).
Before entering into such arrangements, the Society evaluates potential partners and checks diligently to ensure its partners are competent, credible providers of the services they will be offering together with the Society. In looking for potential partners, the Society will often, although not always, focus on providers that are offering new services or novel approaches. While this opens us to criticism that we are merely chasing fads, it is our belief that we have a duty to members to help them stay abreast of the latest developments in the quality field.
The Society is also mindful that in entering a partnership it may be creating the impression that it is favoring one entity over another or that it unfairly is competing with its own members. The question of competition with members can be an intricate one, but it is not new to ASQ or to professional associations in general.
For example, most professional associations are publishers and providers of educational services, and most include in their membership individuals or entities that are publishers and providers of educational services. But in these circumstances, a professional association should not be viewed as a competitor in the traditional sense. Rather, the role of the professional association is to develop and nurture market environments in which many others can also flourish.
We believe that this general statement applies to ASQ's relationship with the Six Sigma Academy. The Society's association with the academy is intended to accomplish two primary goals: to raise the general level of awareness about Six Sigma and to expand access to Six Sigma training for ASQ members. In doing so, the Society is creating an environment in which members can improve career prospects with their current employers or in fact pursue business opportunities in the growing field of Six Sigma consulting. Revenues derived from this partnership go to defray the operating costs of the Society and thus are a benefit to all members.
A final question that needs to be addressed is the editorial independence of Quality Progress. According to its mission statement, the magazine's role is to be "the world's leading source of timely information about quality principles, tools and techniques." Since this mission is similar to the mission of ASQ as a whole, it stands to reason that the magazine will write about subjects in which the Society is involved and perhaps those in which it has a business partnership.
In selecting articles for publication, the magazine's staff is expected to exercise its own independent judgment and as necessary seek input from members of its Editorial Review Board and from the Publications Management Board. This is what happened in this case. No one in ASQ's senior management or in its volunteer leadership structure directed, or even suggested, that Quality Progress publish an article about Mikel Harry or Six Sigma.
The QP mission statement described above goes on to state that a key part of the magazine's role is to "stimulate discussion on achieving performance excellence." If nothing else, as the letters above demonstrate, the recent "Cowboy Quality" article has achieved this element of the mission.
The question of whether Mikel J. Harry should be compared to such quality luminaries as Walter A. Shewhart and W. Edwards Deming is one that will be decided ultimately by the test of time. But those who take issue with the fact that I made the comparison are overlooking the way I framed it.
The comparison I made was written using a tense called the present progressive, which implies an ongoing action not a finished state. This is only fair in that Harry is still in his 40s and will presumably be active in the quality field for many more years.
Furthermore, the sentence made the comparison on the basis of the effect that these men have had on business practices and management thinking, not on overall contributions to society.
When you consider the enthusiastic, nay evangelical, support for Six Sigma that Harry has won from Jack Welch--who is repeatedly cited in polls as the most widely admired executive in business--and when you consider that Six Sigma is being used by an increasing number of world-class companies, it is hard to overestimate or exaggerate the impact Harry is having.