On My Mind
Senior executives often want to know how their organization is performing over time. Has our new strategy improved performance? Are we growing? Is our quality improving? Is customer satisfaction increasing? Do we have a safer place to work? In the Hoshin Kanri process, we want to know if we are moving towards our targets. There are a number of analytical techniques that can provide insight into performance over time. Let’s say that we’ve collected data over time for a performance metric—this is our “time series data.” We can simply plot the data (using most software packages) creating a Time Series Chart. I like to lag the data one time period and then calculate the point-to-point differences and percent changes. Creating these two Time Series Charts gives us insight into growth—or the lack of it. We can create a Statistical Control Chart to help us determine if we have special causes of variation in addition to common causes of variation. If we have a statistical software package like Minitab, then we can build an Autoregressive Integrated Moving Average (ARIMA) model for our time series data to better understand its behavior and generate forecasts. This class of models uses correlational techniques and is well-suited for data that exhibits nonstationary behavior. I will be conducting a public seminar on these techniques June 20. Please join me if you would like to learn more.
Sadly, Dr. George E. P. Box passed away on March 28 at the age of 93. He was one of the world’s leading statisticians for a number of decades having conducted pioneering work on the design of experiments, response surface methodology, evolutionary operation, time series analysis, data transformations, and more. I was fortunate to have spent a lot of time with George during my time at the University of Wisconsin for my doctoral studies. I become a member of the Center for Quality and Productivity Improvement (CQPI), which he led. There were faculty members and students from many disciplines who attended CQPI meetings (Engineering, Business, Statistics, Education Administration, etc.). George also offered his famous “Monday Night Beer and Statistics” seminar. I believe students could earn one credit for the seminar—although I never formally registered. This took place in George’s home and was open to the public. It lived up to the billing as you were guaranteed plenty of beer and statistics. As I recall, the University eventually had George remove the word “beer” from the title of the seminar. I took two independent study courses from George. I met with him one hour each week during the Spring Semester of 1989. We discussed one chapter each week from the book, Empirical Model-Building and Response Surfaces by Box and Draper. We repeated the process the Spring Semester of 1990 only this time I studied the book, Time Series Analysis: Forecasting and Control, by Box and Jenkins. George was brilliant, humble, humorous, and kind. I treasured learning both the technical details of the analytical techniques and George’s stories about how they were developed. He will be greatly missed by many.
I will be attending an event to “Celebrate the Life of Dr. George E. P. Box” at 2:00 p.m. on Sunday, May 5, 2013, at First Unitarian Society, 900 University Bay Drive, Madison, WI. Strategic Improvement Systems, LLC is one of the sponsors for the Performance Excellence Network (PEN) Annual Conference May 14-15, 2013 at the Earle Brown Heritage Center in Brooklyn Center, MN. There is still time to register for my upcoming public seminars: Creating Standard Work (May 23); Leading Design Projects (June 6); and Measuring Performance Over Time (June 20). Brochures are accessible by selecting the “SERVICES” tab on the home page.
To learn more about Dr. Box and his life, you should read the just-published autobiography titled, An Accidental Statistician: The Life and Memories of George E. P. Box.