W.E. Deming

William Edwards Deming was born in Sioux City, Iowa on Oct. 14, 1900, and died in Washington D.C. on Dec. 20, 1993. His work as a statistician is best known in the area of industrial quality control, where his use of statistical methods helped him develop ideas that promoted better quality and more efficient industry.

After getting degrees at the University of Wyoming and University of Colorado, Deming got his Ph.D. in mathematical physics from Yale University in 1928. He then went on to work as a mathematical physicist for the U.S. Department of Agriculture, a statistical advisor for the U.S. Census Bureau, and then from 1946-1993 as a statistics professor at New York University and Columbia University.

It was in the 1930's, while studying under statistician W.A. Shewhart that Deming became interested in the use of statistical analysis to devise quality control and process improvement methods for industry. The ideas he developed were summarized i n his famous 14 Points for management as follows:

Deming's 14 Points

· Create a constancy of purpose for the improvement of product and service. Companies, should invest in innovation, research, and maintenance to make more money in the long run.

· Adopt a new philosophy of eliminating defective products and worker incompetence as it is a drain on revenue and quality.

· Eliminate the use of mass inspection. Quality comes from improving the process, not mass inspection.

· Do not choose suppliers based on price. Consider quality and choose the supplier that uses statistical quality control measures.

· Improve production and service constantly. Always look to improve quality.

· Use modern training methods. All employees should be properly trained.

· Use modern methods of supervision. Supervisors should be active with the employees to help make better products.

· Drive out fear. Employees must feel secure about their job, and not be afraid to report problems to management.

· Create a sense of teamwork among the different departments in the industry.

· Eliminate slogans and goals. It is better to explain what improvements need to be made and let employees set their own goals, so that they will be committed to them.

· Eliminate standards and numerical quotas. Quotas do not take into account quality, and employees will ignore quality to meet the quota.

· Get rid of barriers that do not allow the worker to do their job. Employers should listen to problems and suggestions, and remove equipment that blocks good workmanship.

· Institute education. Management and workers should both be educated in teamwork and statistical quality control techniques so that everyone is more likely to strive for quality and seek ways to improve it.

· Construct a management team that will carry out these points.

It was these methods and ideas that led Japanese business leaders to invite Deming to Japan to teach his new methods in 1950. It is widely regarded that Deming's influence had much to do with Japan's recovery from W.W.II and their economic boo m in the 20th century. In fact, Japan awards the Deming Prize annually to Japanese corporations that excel in quality control. However, despite Deming's success on Japanese industry, his ideas were largely ignored in America until the 1980's when U.S. c orporations needed to compete more effectively with foreign markets.

So what do Deming's ideas for quality control have to do with statistics? First, they were centered on the idea of statistically recording problems, analyzing the causes, correcting them, and then recording the effects on quality. In his 14 p oints, Deming stresses the education of statistical techniques and the frequent use of statistical quality control charts to make everyone aware of the current level of quality. He also says that a statistical expert should be employed by the company to help identify the problem, collect data to help find a solution, and analyze this data to draw inferences.

Other ways in which statistics can be used in industrial quality control include analyzing the quality control of suppliers, determining process capabilities, and analysis of different production procedures. The product can also be analyzed st atistically through accelerated life tests, and survey's can be made to determine customer preferences and satisfaction with quality.

Deming's ideas have influenced the world of industry more than that of statistics, but the use of statistics to formulate his ideas are very evident. His ideas on control of quality have had great impact, particularly in Japan, and will contin ue to influence the processes of American industry in the years to come.