Vol. 12, No. 2
February 2010
PQ Systems
Contents
Tweet with us on TwitterRead our Quality Blog
 

Send Quality eLine
to a friend!

Just type in your friend's email below:

 

Sign up
If you received this newsletter from a friend and want your own subscription to Quality eLine, click below.

Subscribe to Quality eLine

 
Software

 

   

Six Sigma and more:
Dave Schwinn ruminates about numbers

I recently attended a PowerPoint presentation of a statistical study. Some of the numbers were big. Some of the numbers were little. Some comparisons were made. The differences between some numbers were large and some were small. I saw no control charts. I saw no trend lines. I wasn’t sure what mattered and what did not. I was not sure of the purpose of the presentation. I walked away with more questions than answers.

Walking away with questions is not all bad, but I believe that we, as Six Sigma professionals, have the capability and the intent to provide some answers with the numbers. I do not say this to be critical. I say it to remind us what many, if not most, presentations are like. Beyond the hallowed halls of Six Sigma, this is how people behave.

The presentation reminded me of another problem with the numbers. Beyond the Six Sigma community, people tend to gather data mostly because it is easy to gather and, probably, because it was useful at another time. I suspect that, even within the Six Sigma community, some numbers are imposed upon us for these reasons. These “easy” numbers are not all bad. They provide historical perspective to the extent that we use them for analysis. Because they have been around for a while, we know how to manage them.

Here lies a risk. The people who have been most successful in our organizations are probably the ones who have most successfully managed the existing data. Therefore, they like those metrics and are reluctant to have them change. I’m reminded how Ford’s retired president and chairman of the board, Don Petersen, would introduce Dr. W. Edwards Deming to the management team, when Ford was attempting to embrace Deming’s philosophy. He would describe the need for deep change by pointing out that many of the people in the room had been promoted for all the wrong reasons. The essence of Dr. Deming’s philosophy, of course, centered on understanding numbers using analytic studies.

Deming differentiated between enumerative statistical studies, such as descriptive and inferential statistics, from analytic statistical studies such as control charts. He pointed out that enumerative studies are intended to help understand what exists. The Census, for example, estimates how many people populate the United States. Sampling a shipment of incoming product can estimate how many products are defective within that shipment. Enumerative studies help you make decisions about what exists, such as deciding whether or not to reject the product shipment.

Analytic studies help you predict the future. If, for example, you want to have some sense of what the next shipment will be like, enumerative studies are not useful. Analytic studies are. This distinction is important because most decisions made by managers are intended to influence the future. Having some knowledge of what is likely to happen in the future if the manager does nothing is probably not a bad idea…duh! Being able to differentiate between common causes and special causes of variation is another very valuable management decision-making tool. Analytic studies, of course, also provide that bonus.

Despite all this and despite the emotion-filled confessions that academic statisticians used to make to Deming at his Seminars for Statisticians, most statistics course still seem to focus on enumerative studies. As a result, perhaps, most managers apparently don’t understand the power that analytic studies have for improving their decision-making power. Unless they have a strong background in the Deming philosophy or in Six Sigma, your managers also probably don’t understand. They also probably don’t understand that how they manage the numbers, even if they are basing them on analytic studies, may be detrimental to their intent.

Deming also taught us about system sub optimization. Myron Tribus gave us the Perversity Principle. Alfie Kohn helped us understand the potentially dangerous results of artificially created win-lose competition. Understanding these principles in the management of the numbers is useful.

Jim Bakken, a former vice president at Ford, noticed many years ago that our focus on short-term profits left no time for preventive maintenance during the high points of our sales cycles and no money for preventative maintenance during the low points of our sales cycles. Hence, no preventative maintenance. His observation led to a good opportunity for improvement that Jim led.

Mike Cleary, the founder of PQ Systems, thought he’d try out these new understandings and found out years ago that he could eliminate the internal win-lose competition perceived by the PQ sales folks with his traditional sales incentive system. He changed it and most everything about PQ sales improved.

Not understanding these principles is a much more common management practice. When financial institutions focus entirely on showing short term profits, they sometimes disregard risk, and financial systems fail. When publicly-held companies focus on the next quarter’s profits and on the artificially created profit goals set by Wall Street, long-term viability goes down the tubes. When nonprofits think they are competing for a fixed sum of money from their funder, they do things that are inconsistent with their mission and that they don’t know how to do anyway. That’s one of the most common ways nonprofits waste money.

All is not lost. You understand the numbers. Even if the people you work for do not, there are some things you can do. Here goes:

  • Start with the mission of your organization. Gather numbers that tell you how well you’re accomplishing that mission. You may already be doing that, but remember Deming’s observation that the most important numbers are unknown and unknowable. It’s hard to measure “success” as a mission, for example, but you can establish a set of surrogate measures. Don’t accept metrics that are simply easy to gather.
  • Gather the data in a way that maximizes the power of analytic studies.
  • Use the power of control chart theory.
  • Remember the danger of rewards, punishments, and internal win-lose competition.
  • Remember the need for soft skills. Over the years, I continue to notice that people interested in improvement tend to focus either on the numbers side or on the people side. You need both! Either continue to develop yourself to embrace both sides or get a partner who can provide the balance you think you need.

Here is one final recommendation. If doing these things seems beyond your control, just control what you can. You can even go beyond by asking permission. You can do an experiment in a new way of operating. You can even do the experiment in secret and show the results afterward. This is all risky business. Change is risky business. Please try. There’s a lot of work to do out there. We can all make a difference.

As always, I welcome your comments and questions. I’m at support@pqsystems.com.

 

Copyright 2010 PQ Systems.
Please direct questions or problems regarding this web site to the Webmaster.