Q: I work for one of the largest service delivery organizations in the federal government, providing benefits to almost 55 million people. I have been asked to explore how to measure the total cost of quality (TCOQ) in the government or service sectors. There is a lot of literature about the formula, such as TCOQ = prevention costs + appraisal costs + failure (external and internal), the categories and elements. However, I have not found it applied in the government, healthcare or service sectors. Rather, I get examples such as Six Sigma, lean Sigma, ISO 9000 and the Baldrige award. Are there any local, state or federal governmental organizations (or any sector besides manufacturing) that have attempted to or are currently applying TCOQ methodology, or costing out the four categories?
Carolyn R. Neuwirth
Senior management analyst
Office of quality performance,
division of management analysis
Social Security Administration
A: You have been given a big challenge with an equally big opportunity. Because you work for a very large government agency that does not have a cost of quality program, the opportunity is likely enormous.
You mentioned that you found examples like Six Sigma, lean Sigma and ISO 9000. COQ is part of the body of knowledge for ASQ certified Black Belts. Therefore, agencies that are implementing Six Sigma programs might be starting COQ programs. You might uncover helpful information if you pursue these leads a little further. Implementation of traditional COQ programs within governmental agencies is spotty. There is evidence of implementation at various local facilities such as naval shipyards. Perhaps the most successful implementation is at NASA. A formal COQ program was in place as early as 1964. More recent initiatives include highly sophisticated measures to reduce the COQ in software development.
After reading your question, my hunch is that you are looking for guidance from someone who has already implemented a successful program. His or her insights might help you gain management support and overcome bureaucratic obstacles. However, the nuts and bolts of the program will probably be unique to each agency. After all, quality problems at the Internal Revenue Service are quite different from mistakes within the Social Security Administration.
I suggest you dive in and get started. An excellent source for information is Principles of Quality Costs: Principles, Implementation, and Use, third edition, by ASQ's Quality Costs Committee. Published in 1999, it is available from ASQ's Quality Press. You might also want to contact someone on the committee. Some of its members might have first-hand experience implementing a program within a government agency.
It might behoove you to start small. Pick a particular problem, apply the principles, and start collecting data. Your agency serves 55 million people, so it probably processes millions of forms per year. If the form is poorly designed, your agency could spend hundreds of thousands of hours per year processing (or rejecting) forms with errors and omissions. The cost of redesigning the form, printing new forms and retraining your agents will not be trivial, but the cost of perpetuating the errors for years to come is far greater.
Studies have shown that the cost of an external failure can be as much as 100 times the cost of preventing the problem in the first place. As your successes mount, your agency will naturally shift a greater percentage of the budget from correction to prevention, and total costs will decrease. Just how big is this opportunity? Some analysts suggest that the cost of poor quality increases total organizational costs by as much as 20%.
For more information:
"Implementing NASA's Quality Program," Howard M. Weiss, Proceedings of 10th National Symposium on Reliability and Quality Control, January 7-9, 1964, p. 57.
"Developing Risk-Based Financial Analysis Tools and Techniques to Aid IV & V Decision Making," Nancy Eickelman, 2001. www.nasa.gov/centers/ivv/ppt/172489main_Nancy_Eickelmann_Motorola.ppt (case sensitive).
Master Black Belt, principle consultant Quintiles Consulting
Q: I have been considering an application for a cumulative sum (CUSUM) chart. I would like to better understand the reason for the slack factor in the CUSUM statistic. Instead of simply using a cumulative deviation from target, the CUSUM increases or decreases the target by some slack factor (a multiple of the standard deviation). When plotting the hi CUSUM, a higher than actual target is used. When plotting the lo CUSUM, a lower than actual target is used. Why is the slack factor necessary?
Quality manager, 3M Co.
A: The slack factor in a CUSUM chart is not necessary—it is an option. It is used if there is an amount of drift in a process that is considered acceptable. Using a slack factor will keep the chart from giving a signal when you really don't need to react.
For example, if you have a very capable process, one with a high Cp, you might want to allow the mean of the process to shift some from the nominal value. You know, based on your capability studies, there is a very low chance of producing any product outside of specifications even when the process is not centered (your Cpk is still high). So you use a slack factor in the CUSUM calculations to prevent the sum from building and showing a signal. You avoid the cost of stopping a process that is producing product acceptable to the customer.
An alternative to CUSUM charting with slack factors is using acceptance control charts (sometimes called modified control charts). With this approach, control limits are calculated inward from the product specification limits, rather than outward from the nominal value. The distance is a function of known short-term process variation. Again, this allows the process mean to shift by an amount that is considered acceptable.
For More Information:
Tabular CUSUM Charts at www.micquality.com/six_sigma_glossary/control_charts.htm#13.
CUSUM and EWMA Control Charts at www2.isye.gatech.edu/~jbanks/coursefiles/QCpptFull/Chap08Rev.ppt (case sensitive).
"Building an Acceptance Chart," www.qualitydigest.com/june00/html/accept.html
Peter E. Pylipow
Senior design excellence engineer
Vistakon–Johnson and Johnson Vision Care
Q: We are writing our Texas Award for Performance Excellence application. What are some benchmarks regarding training hours for employees?
Director of HR
A: When asked if there is a standard for training hours, yes, there are historical and anecdotal cases in some companies' quality and improvement journeys' statements of hours. But this leads many to believe there might be some magic number of training hours that will bring about improvement. There is no such thing. The training must focus on building skills and closing the gap from what is to what is needed based on a full analysis of the plan.
The path suggested is as follows:
- Conduct a job breakdown analysis (JBA). Determine what skills are needed to meet or exceed the necessary requirements for a specific job.
- Do a needs analysis against the JBA to identify the needs or gaps from the ideal. >
- Plan and deliver the training needed.
- Assess the effectiveness of the training through post testing or analysis of improvements in performance.
- Do it again in a year or two as job responsibilities shift around because technology evolves and changes roles and because people move around the organization, which changes the JBA skills analysis.
Many companies, even when doing this type of analysis, want a baseline on overall training. There are two levels:
- Communicate or get all personnel on the same path with overview classes or introductions to various subjects, such as the legal and statutory training that many of us are required to give (right to know, blood-borne pathogens and introductions to company and culture, for example).
- The better training option helps employees attain a level of skill needed to benefit the employee's personal success and his or her contributions to the company. It is not an easy path forward. It takes time to do an analysis and identify the real needs. But this is where improvement comes from—the planned, intelligent delivery of skills needed in today's workplace.
John O. Brown
TU..V Rheinland of North America
Senior auditor, ISO 9000:2000
and TS 16949:2002