Featured Video
This Week in Quality Digest Live
Quality Insider Features
Laurel Thoennes @ Quality Digest
There’s a better way to end this muda
Context plays a critical factor in why users click on a phishing email
Sally Davies
How to style a good doctor
Christina Fattore
For consumers, other things matter more than where something is made
Jama Software
Teams must be able to access, collaborate, update, and test each requirement through to project completion

More Features

Quality Insider News
Replace mechanical indicating applications in smallest AGD size specification class
A new path for local hardware connectivity
mCaliper transfers, processes, and visualizes measurement data collected with tools like digital calipers, micrometers
Robot-served vision, new force-testing products, and electronic gauges at booth No. 135532
Provides accurate visual representations of the plan-do-study-act cycle
Machine demos, technology previews, and a daily happy hour at booth No. 338319
Marking floors is an easy and efficient way to direct behavior, promote safety, and reinforce workplace standards
Other exhibits will feature machine tools to demonstrate tool setting, probing, machine monitoring, and robotic gauging

More News

Matthew Barsalou

Quality Insider

Statistics for the Rest of Us

R.I.P. George E. P. Box, 1919–2013

Published: Monday, April 8, 2013 - 10:47

Thursday, March 28, 2013, was not a good day for the field of quality. On that day we quality practitioners lost the great statistician George E. P. Box, who died at age 93.

I’m not a statistician; however, as a quality professional, I am somebody who needs to use statistics for practical, real-world uses. I believe it is people like me who benefit the most from Box’s many contributions to the field of quality. Box was very involved with educating people in practical statistics. In An Accidental Statistician” he wrote that he realized that “students were learning a great deal about statistical theory but very little on how to use it,” so he invited students and faculty into his home to discuss statistics. For those of us who never had a chance to meet him in person, there are many of his publications available.

Box’s contributions to the field of quality are many. No quality engineer should be without a copy of Statistics for Experimenters: Design, Innovation, and Discovery  (Wiley-Interscience, 2nd ed. 2005), written by Box together with J. Stuart Hunter and William G. Hunter. Box also gave us response surface methodology and popularized two-level factorial designs. He gave us the Box-Cox transformation, which is a type of power transform for transforming data without losing their rank order. Another of his many statistical contributions to quality improvement is evolutionary operation (EVOP), which uses on-line experiments with minor variations in factors and levels to experiment on material that will eventually be delivered to the customer. Together with Gwilym M. Jenkins and Gregory C. Reinsel, Box wrote the classic book, Time Series Analysis: Forecasting and Control  (Wiley, 4th ed. 2008), which is often used by both engineers and financial analysts to “obtain an understanding of the underlying forces and structure that produced the observed data” as well as to perform modeling and forecasting.

Box was more than just a statistician; he also provided nonstatistical guidance on quality improvement. For example, he went beyond statistical number-crunching and recommended using the data from a system to improve it, and he reminded us that quality improvement is everybody’s responsibility. In the article “When Murphy Speaks—Listen” (Quality Progress. Oct. 1989) he also told us that input from the people closest to the system is needed for quality improvement. Good nonstatistical advice from a great statistician.

Box also gave the field of quality the concept of iterative experimentation, which is useful for both statisticians and nonstatistical quality professionals. In iterative experimentation, an investigator moves from conjecture to experimentation and then, based on the experimental results, back to conjecture to conceptualize a new experiment. The “d” in figure one represents the process of analyzing a hypothesis and constructing an experiment, and the “a” represents the analysis of the experimental results. The iterative process is repeated over and over as each cycle generates information that can lead to new knowledge. This can be used in root cause analysis or process optimization. In Statistics for Experimenters, iterative experimentation is referred to as the “iterative inductive-deductive process,” and it should be required reading for everybody who performs experiments in industry.

Figure 1: Iterative experimentation

As Box points out in his memoir An Accidental Statistician: The Life and Memories of George E. P. Box  (Wiley, 2013) he had not intended to be a statistician. He had almost completed a degree in chemistry but left the university to join the British army at the start of the World War II. He was assigned to an experimental station that was looking for actions it could take if poison gas was used during the war. Box asked the head of the experimental station for a statistician and was told he would need to do the statics himself because he had once tried to read a statics book by Ronald A. Fisher. Box then read many books on statistics and eventually went to meet Fisher to discuss a statistical problem. The army had no way to officially dispatch a sergeant such as Box to see a professor, so his official reason for travel was “to take a horse to Cambridge.”

After the war, Box studied statistics under the statistician Egon S. Pearson. Box then worked for Imperial Chemicals, where he used statistics to improve the company’s processes. In 1956 Box went to Princeton University at the request of John Tukey.

Box would eventually start teaching at the University of Wisconsin-Madison’s Department of Industrial and Systems Engineering. He was also involved with the University of Wisconson-Madison’s Center for Quality and Product Improvement, where many of Box’s papers can be read.

Box received many awards during his long career, including the Royal Statistical Society’s Guy medal in Silver in 1964, the American Statistical Society’s Wilks Award in 1972, and the Royal Statistical Society’s Guy medal in Gold in 1993. He also won the American Society for Quality’s Shewhart Medal in 1968 as well as a Brumbaugh Award, the Youden Prize from the ASQ’s Chemical and Process Industries Division, and the Deming Medal from the ASQ’s Metropolitan Section. In addition to earning many statistics-related prizes and awards, there is also a medal named for Box. The Box Medal for Outstanding Contributions to Industrial Statistics is awarded by the European Network for Business and Industrial Statistics for “the application of statistical methods in European business and industry.”

We can at least consider ourselves fortunate that Box published his memoir. Although he has left us, we can safely assume his memoir will contain many new lessons that he has left behind for us to ponder.


About The Author

Matthew Barsalou’s picture

Matthew Barsalou

Matthew Barsalou is a statistical problem resolution master black belt at BorgWarner Turbo Systems Engineering GmbH. He is an ASQ-certified Six Sigma Black Belt, quality engineer, and quality technician; a TÜV-certified quality manager, quality management representative, and quality auditor; and a Smarter Solutions-certified lean Six Sigma Master Black Belt. He has a bachelor’s degree in industrial sciences, and master’s degrees in engineering, business administration, and liberal studies with emphasis in international business. Barsalou is author of Root Cause Analysis, Statistics for Six Sigma Black Belts, The ASQ Pocket Guide to Statistics for Six Sigma Black Belts, and The Quality Improvement Field Guide.


Thank you, Matthew

Thank you, Matthew, for that tribute. George Box was certainly one of the most original thinkers in statistics. One of his more revolutionary observations was that "the domination of Statistics by Mathematics rather than by science has greatly reduced the value and the status of the subject." This quote, from the abstract of "Scientific Statistics, Teaching, Learning and the Computer," discussed the value of statistics in an iterative learning paradigm,  very different from the theorem-proof paradigm of mathematics. He noted: "An important issue of the 1930's was whether statistics was to be treated as a branch of Science or of Mathematics. To my mind unfortunately, the latter view has been adopted in the United States and in many other countries. Statistics has for some time been categorized as one of the Mathematical Sciences and this view has dominated university teaching, promotion, tenure of faculty, the distribution of grants by funding agencies and the characteristics of statistical journals. All this has, I believe, greatly limited the value and distorted the development of our subject."

There are too few influential statisticians who think that way. The loss of one of the most influential thought leaders in our subject will be keenly felt. This paper, and the first chapter of Statistic for Experimenters, should be required reading at the beginning of every statistics course.


Box, G.E.P. (June, 1996). Scientific statistics, teaching, learning and the computer. CQPI Report 146, 1-7.

Box, G.E.P., Hunter, J. S., & Hunter, W. G., (2005). Statistics for experimenters. (2nd Ed.) Hoboken, New Jersey: Wiley-Interscience.



I'm no Statistics fan: if there's still some freedom in this world, one could choose the music he wants to listen to. And I don't like many Statisticians' music, just as I don't like Rap music. Statistics started as a prediction-based tool but it was made to grow to a self-feeding beast, that's now - I sincerely hope - decaying to a saurus. I've seen too many tons of charts been idiotically filled by line operators, and left to rot in the quality manager's office: yes, "Statistics for the Rest of Us". The Ultimate Rest. Thank you.