The revolution of six-sigma: an analysis of its theory and application.

Drake, Dominique ; Sutterfield, J.S. ; Ngassam, Christopher 等

INTRODUCTION

Data is good but good data is better. However, in order to get this good data it is important to have measures in place such as quality control systems to ensure information accuracy. Quality has evolved over the past two centuries when it first became an important business measure of comparison. The ways in which it has been used to define and assess quality have also evolved as new business practices and degrees of acceptance are enforced. One of the first definitions of quality was conformance to valid customer requirements or the Goalpost View, Gitlow & Levine and Levine (2005). As long as the company's process met the customer's requirements, then both the customer and the company were satisfied. Frederic Taylor in his theories of scientific management subscribed to this viewpoint. He believed segmenting jobs into specific works tasks was the best way to control outputs because then efficiency would increase, Basu (2001). Employees were told what to do and how to do it with little or no input. They also rarely worked in teams or cross-functional efforts because each person had his or her own tasks to complete. Moreover, postproduction inspection was the primary means of quality control, which showed the lack of concern for waste and error at intermediary steps in the process, Evans and Lindsay (2005). Under this traditional approach, companies did little to understand customer requirements, neither external nor internal, and rarely focused efforts on finding a way to improve quality if it did not come from a technological breakthrough.

As companies began to realize this approach did not take into consideration the cost of waste and variation, many began looking for new ways to measure and define quality. This led to a more modern definition of quality being accepted as the predictable degree of uniformity and dependability, Gitlow & Levine (2005). This Continuous Improvement View supported the notion that every process contains an element that can be improved. Inherently, continuous quality improvement focused on fine tuning parts of the whole through incremental changes. When a problem was discovered, it was addressed until the next problem was discovered and so on. Industrial scientists such as William Edwards Deming, Joseph Juran, Walter Shewhart, Harold Dodge and others were at the forefront of this era. They began by applying statistical methods of control to quality. These methods showed the shift from inspection of the final product and putting a band-aid on the problem to actually implementing quality control into the manufacturing process. This also illustrated the increasing importance of management decision-making in the quality control efforts instead of simply finding a quick fix to the problem so it fell within the specification limits of product or service satisfaction and acceptance.

Deming and Juran believed in a new concept called Total Quality Management. Research by Evans and Lindsay (2005) indicate this viewpoint was based on three fundamental principles:

1. Focus on customers and stakeholders

2. Participation and teamwork by everyone in the organization

3. A process focus supported by continuous improvement and learning

Through Total Quality Management (TQM), workers were empowered to provide input throughout the entire process to ensure and instill confidence that products met customer specifications. In meeting the customer requirements, the term quality assurance became widely used as any planned and systematic activity directed toward providing consumers with products of appropriate quality along with the confidence that products meet consumers' requirements Evans and Lindsay (2004). This extension of the Continuous Improvement viewpoint coined the terms Big Q and Little Q, which referred to the quality of management and the management of quality, respectively Evans and Lindsay (2004). Driven by upper management, TQM focused on not just identifying and eliminating problems that caused defects but educating the workforce to improve overall quality. Upper management realized it needed to focus more on identifying and eliminating problems that were causing the defects throughout the process rather than just waiting until the end for post-production inspections. In order to do this, continuous education programs were implemented at all levels. As time passed, these total quality management efforts soon began to fall short of expectations of businesses in the United States. Companies vowed they could promise a certain level of quality by getting it right the first time, but in the end the processes were still not meeting all needs and requirements of the consumer. TQM was joined by many other acronyms like JIT, MRP and TPM that all guaranteed an unreachable quality standard Basu (2001). Because these methods focused so greatly on the parts of the whole rather than the whole, they fell short in implementing rapid changes. This led to the introduction of certain holistic quality control programs, one of the most recent being Six- sigma.

Six-sigma is a discipline based on the Greek letter sigma which is the conventional symbol for standard deviation. Six-sigma is a very structured process that focuses on quantitative data to drive decisions and reduce defects Benedetto (2003). Many companies have tried to implement this program without truly understanding the theory and concept behind it. Because of such misguided efforts, Six-sigma has received a somewhat jaded reputation, but it is not the discipline that is flawed but rather the application of it. As Evans and Lindsay note, a cookbook approach in which management reads the latest self-help book advocated by business consultants and blindly follows the author's recommendations, will destine Six-sigma or any other management practice to failure because experience only describes what happened and is no help to the emulating management team Evans and Lindsay (2004). Understanding the subtle theory, on the other hand, will enable one to better understand the cause and effect relationships that are applied in Six-sigma and other management practices. In this paper, we will further develop the historical roots of the quality revolution already illustrated, show how quality revolution developed into Six-sigma, delve further into the underlying theory of Six-sigma and then analyze the uses of some Six-sigma tools used in an effective, coherent Six-sigma program.

LITERATURE REVIEW

Multiple scientists have contributed to the evolution of quality management. The main three who added to the development of six-sigma and will be discussed here are William Edwards Deming, Joseph Juran and Philip Crosby. Deming had a Ph. D. in physics and was trained as a statistician at Yale University. He and Juran both worked for Western Electric during the 1920s and 1930s Evans and Lindsay (2004). From there he started working with the U.S. Census Bureau where he perfected his skills in statistical quality control methods. He opened his own practice in 1941 and thus began teaching SQC methods to engineers and inspectors in multiple industries. Deming, better known as the Father of Quality, was widely ignored in the United States for his works. As a result, he went to Japan right after World War II to begin teaching them SQC. His work there is what truly bolstered the "Made in Japan" label to its well-respected level of quality today W. Edwards Deming (2007). Dr. Deming's most famous work is the 14 points in his System of Profound Knowledge. There are four interrelated parts in the program: Appreciation for a System, Understanding of Variation, Theory of Knowledge and Psychology, Evans and Lindsay (2004). In the first point, Deming believed it was poor management to purchase materials at the lowest price or minimize the cost of manufacturing if it were at the expense of the product. The second and third points emphasized the importance of management understanding the project first and foremost before taking any steps to reduce variation. The methods to reduce variation include technology, process design and training. Under the fourth point, Deming stressed understanding the dynamics of interpersonal relationships because everyone learns in different ways and speeds and thus the system should be managed accordingly. Other points in the Deming school of thought were on-the-job training, creating trust, building quality into a product or service from the beginning and inclusion of everyone in the company to accomplish project improvement. The common methodology used by Deming for improving processes is PDCA which stands for Plan--Do--Check--Act.

Joseph Juran is well known for his book, the Quality Control Handbook published in 1951. He too went to Japan in the 1950s after working at Western Electric to teach the principles of quality management as they worked to improve their economy. His school of thought was different from Deming's in that he believed to improve quality companies should use systems with which managers are already familiar. When consulting firms, he would design programs that fit the company's existing strategic efforts to ensure minimal rejection by staff. The main points he taught were known as the Quality Trilogy: Quality Planning, Quality Control and Quality Improvement. Quality planning was the process of preparing companies to meet quality goals by identifying the customers and their needs. Quality control was the process of meeting quality goals during operations with minimal inspection. Quality improvement was the process of breaking through to unprecedented levels of performance to produce the product (Evans and Lindsay, 2004). His philosophy required a commitment from top management to implement the quality control techniques and emphasized training initiatives for all.

The third philosopher who contributed greatly to six-sigma was Philip Crosby. Crosby worked as the Corporate VP for Quality at International Telephone and Telegraph for 14 years. His most well-known work is the book Quality is Free where he emphasized management should "do it right the first time" and popularized the idea of the "cost of poor quality." Crosby emphasized that management must first set requirements and those are on which the degree of quality will be judged. He also believed doing things right the first time to prevent defects would always be cheaper than fixing problems which developed into the cost of poor quality idea. Similar to six-sigma, Crosby focused on zero defects and created four Absolutes of Quality Management to ensure that goal was accomplished Skymark Corporation (2007):

1. Quality is defined as conformance to requirements, not as "goodness" or "elegance"

2. The system for causing quality is prevention, not appraisal

3. The measurement standard must be zero defects, not that's close enough

4. The measurement of quality is the Price of Nonconformance, not indices.

These philosophies have been categorized as more evolutionary because the changes resulting from their implementation were more gradual. As time continued, managers became increasingly disappointed with TQM and searched for philosophies that would create more drastic changes in the existing processes and systems. These revolutionary processes recognized that processes are key to quality, that most processes are poorly designed and implemented, that overall success is sensitive to individual sub-processes success rates and that rapid, dramatic change requires looking at the entire process. One of the first of these revolutionary philosophies was Reengineering. Reengineering is defined as the fundamental rethinking and radical redesign of business processes to achieve dramatic improvements in critical, contemporary measures of performance Benedetto (2003). It places heavy reliance on statistics and quantitative analysis.

Six-sigma emerged during this revolutionary time period after reengineering. The two approaches are very similar but also have many differences. Whereas reengineering focuses on the statistical aspect, Six-sigma places more emphasis on data to drive decisions. When comparing six-sigma to Total Quality Management it is very apparent that great changes have come about in the evolution of quality. Six-sigma has a more structured and rigorous training development program for those professionals using it. Six-sigma is a program owned by the business leaders, also known as champions, while TQM programs are based on worker empowerment. Six-sigma is cross functional and requires a verifiable return on investment in contrast to TQM's function or process based methodology that has little financial accountability Evans and Lindsay (2004).

Six-sigma started as a problem solving approach to reduce variation in a product and manufacturing environment. It has since grown to be used in a multitude of other industries and business areas such as service, healthcare and research and development. It represents a structured thought process that begins with first thoroughly understanding the requirements before proceeding or taking any action. Those requirements define the deliverables to be produced and the tasks to produce those deliverables which in turn illustrate the tools to be used to complete the tasks and produce the deliverables Hambleton (2007). Six-sigma was initiated in the 1980s by Motorola. The company was looking to focus its quality efforts on reducing manufacturing defects by tenfold within a five-year period. This goal was then revised to a tenfold improvement every two years, 100-fold every four years and 3.4 defects per million in five years Gitlow & Levine (2005). A defect in six-sigma terms is any factor that interferes with profitability, cash flow or meeting the customers' needs and expectations. This led to the next important concept of six-sigma: critical to quality. Critical-to-Quality (CTQ) items are points of importance to the customer be it internally or externally, Adomitis and Samuels (2003). Six-sigma projects aim to improve CTQs because research has proven that a high number of CTQs correlates with lost customers and reduced profitability. Ultimately, the company will have reduced the number of defects and focused on the CTQs such that it will cost more to correct the defects any further than to prevent their occurrence in the first place.

This thought was advanced in the 1980s by Dr. Genichi Taguchi with his Quality Loss function. Prior to that time, it was thought that the customer would accept anything within the tolerance range and that product costs do not depend on the actual amount of variation as long as it was still within the tolerance range, Evans and Lindsay (2004). Taguchi introduced a model that discredited that school of thought by showing the approximate financial loss for any deviation from a specified target value--an idea that coincides with six-sigma's inclusion of financial data in the assessment of a process or function. His work contradicted the original concept and showed that variation is actually a measure of poor quality such that the smaller the variation is about the nominal level, the better the quality is. As the graph below depicts, this loss function also applied an economic value to variation and defects from the standard. As the process or project deviates from its target value, costs become increasingly higher and the customer becomes increasingly dissatisfied. Quality loss can be minimized by reducing the number of deviations in performance or by increasing the product's reliability. In six-sigma, the goal is to minimize the number of defects and keep the process as close to the nominal value as possible to avoid high degrees of variability. The concepts of Taguchi are part of Six-sigma's key philosophy of reducing defects and variation from the norm.

[FIGURE 1 OMITTED]

As previously mentioned six-sigma is a product of the statistical methods era of controlling quality. It stressed the common measure of quality, the defect. In measuring output quality, the key metrics used are defects per million opportunities (DPMO), Cost of Poor Quality (COPQ) and sigma level, Druley and Rudisill (2004). Defects per million opportunities takes into consideration the number of defects produced out of the entire batch and the number of opportunities for error. It is important to assess quality performance using the defects per million opportunities because two different processes may have significantly different numbers of opportunities for error, thus making comparison unbalanced. Furthermore, large samples provide for a higher probability of detection of changes in process characteristics than a smaller sample. When Motorola created this ultimate goal of six-sigma, it was set to be equivalent to a defect level of 3.4 defects per million opportunities. According to Evans and Lindsay, this figure was chosen because field failure data suggested that Motorola's processes drifted by this amount on average, Evans and Lindsay (2004). The cost of poor quality metric was introduced by Crosby. It includes the costs of lost opportunities, lost resources and all the rework, labor and materials expended up to the point of rejection. This metric is used to justify the beginning of a six-sigma project. When a company is choosing which project to undertake, the project with the highest COPQ will most likely be selected. Lastly, the sigma level indicates the degree of quality for the project. Sigma is a statistical term that measures how much a process varies from the nominal or perfect target value based on the number of DPMO. As mentioned, Motorola's goal was to have 3.4 defects per million opportunities which would happen at the six-sigma level. The DPMOs for various sigma levels are compared in Table 1 below (iSixSigma, 2002).

One phenomenon that has been widely discussed in relation to six-sigma is the 1.5 sigma shift. Because six-sigma focuses on variation, a chart to understand deviation from the mean should be used to further understand this concept. A Z-table shows the standard deviation from the mean. A process' normal variation, defined as process width, is +/- 3-sigma about the mean. In looking at Table 2, it is clear that a quality level of 3-sigma actually corresponds with 2,700 defects per million opportunities. This number may sound small but it is similar to 22,000 checks deducted from the wrong bank account each hour or 500 incorrect surgical operations each week. This level of operating efficiency was deemed unacceptable by Motorola hence the search for a design that would ensure greater quality levels. A six-sigma design would have no more than 3.4 DPMO even if shifted +/-1.5 sigma from the mean. In looking once again at Table 2, it is evident a quality level of six-sigma actually corresponds with two defects per billion opportunities while the more commonly known value of 3.4 defects per million opportunities is found at the six-sigma level with a +/-1.5 sigma shift off the center value or +/-4.5-sigma. This shift is a result of what is called the Long Term Dynamic Mean Variation, Swinney (2007). In simpler terms, no process is maintained perfectly at center at all times, thus allowing drifting over time. The allowance of a shift in the distribution is important because it takes into consideration both short and long-term factors that could affect a project such as unexpected errors or movement. Also worth noting is that the goal of 3.4 DPMO can be attained at 5-sigma with a +/-0.5-sigma shift or at 5.5-sigma with a +/-1-sigma shift. However as Taguchi pointed out, it is important to minimize the noise factors that could shift the process dramatically from the nominal value. Table 2 below shows the DPMO quality levels achieved for various combinations of off-centering and multiples of sigma, Evans and Lindsay (2004).

METHODOLOGY

Within six-sigma there are different project based methods that can be followed. Among them are Lean Sigma, Design for Six-sigma, Six-sigma for Marketing and the most common DMAIC--Define, Measure, Analyze, Improve and Control, Hambleton (2007). Lean Sigma is a more refined version of the standard Six-sigma program that streamlines processes to its essential value-adding activities. In doing this, the company aims to do things right the first time with minimum to no waste. Wait time, transportation, excess inventory and overproduction are among the wasteful activities that can be eliminated. One test used to identify the value-add activities that will reduce waste is the 3C Litmus Test: Change, Customer, Correct, Hambleton (2007). If the activity changes the product or service, if the customer cares about the outcome of the activity and/or if the activity is executed and produced correctly the first time then it is a value-add activity and should be kept. Lean Sigma was popularized in Japan after World War II by Toyota. The Toyota Production System challenged Ford's standard for a large lot production system by using systematically and continuously reducing waste through small lot productions only. Design for Six-sigma (DFSS) expands Six-sigma by taking a preventative approach by designing quality into the product or process. DFSS focuses on growth through product development, obtaining new customers and expanding sales into the current customer base, Hambleton (2007). It focuses on not just customer satisfaction but customer success because from the inception of the product or service the company will focus on optimizing CTQs. In order to do this, companies use multivariable optimization models, design of experiments, probabalistic simulation techniques, failure mode and effects analysis and other statistical analysis techniques. This method is usually begun only after the Six-sigma program has been implemented effectively. The Six-sigma for Marketing methodology simply means applying six-sigma ideas and principles to improving other functions of the business such as marketing, finance and advertising. The most common methodology is DMAIC. One goal of using the DMAIC methodology is to identify the root cause of the problem and select the optimal level of the CTQs to best drive the desired output. Another goal is to improve PFQT: Productivity, Financial, Quality and Time spent, Hambleton (2007). It is used as an iterative method to combat variation. With the built in ability to contrast variation, DMAIC intrinsically allows for flexibility because as knowledge is learned thorough implementation, assumptions of the root cause may be disproved, requiring the team to modify or revisit alternative possibilities. Kaoru Ishikawa promoted seven basic tools that can be used in assessing quality control. The list has since been expanded upon to include many other tools but the original seven were Cause-and Effect diagrams, Check sheets, Control charts, Histograms, Pareto charts, Scatter diagrams and Stratification, Hambleton (2007). Some of these tools will be further explained below.

In the first step Define, the company clearly defines the problem. It begins the trust and dedication from all stakeholders and all persons included in the project team. The project team consists of the Champion, Master Black Belt, Black Belt, Green Belt and Team Members. Key activities occurring in this phase include selecting the team members and their roles, developing the problem statement, goals and benefits and developing the milestones and high level process map iSixSigma (2007). One tool used in this step is the process flowchart. Four main types of flow charts are considered: top-down, detailed, work flow diagram and deployment, Kelly and Sutterfield (2006). This tool shows how various steps in a process work together to achieve the ultimate goal. Because it is a pictorial view, a flow chart can be applied to fit practically any need. The process map allows the user to gain an understanding of the process and where potential waste or bottlenecks could occur. It also could be used to design the future or desired process. An example process map for a call answering operation is shown below in Figure 2.

[FIGURE 2 OMITTED]

In the second step Measure, the company quantifies the problem by understanding the current performance and collecting necessary data to improve all CTQs. Key activities occurring in this phase include defining the defect, opportunity, unit and cost metrics, collecting the data, determining the process capability. One tool used in this step is the SIPOC Diagram. SIPOC stands for Suppliers, Inputs, Processes, Outputs and Customers. This tool is applied to identify all related aspects of the project--who are the true customers, what are their requirements, who supplies inputs to the process--at a high level before the project even begins (iSixSigma, 2007). These diagrams are very similar to process maps which are also a tool applied at this phase of the DMAIC cycle. A SIPOC diagram is shown below in Figure 3, Simon (2007).

[FIGURE 3 OMITTED]

In the third step Analyze, the root cause of the project's problem is investigated to find out why the defects and variation are occurring. Through this detailed research, the project team can begin to find the areas for improvement. Key activities in this phase are identifying value and non-value added activities and determining the vital few or critical-to-quality elements. The care applied to this phase of the Six-sigma project is very important to the project's success because a lack of understanding and thorough analysis is what causes most defects and variation. Evans and Lindsay (2005) described that it could also be caused by:

* Lack of control over the materials and equipment used

* Lack of training

* Poor instrument calibration

* Inadequate environmental characteristics;

* Hasty design of parts and assemblies.

Multiple tools exist to guide the efforts of this process. They include but are not limited to scatter diagrams, Pareto charts, histograms, regression analyses and fishbone diagrams. A scatter diagram shows the relationship between two variables. The correlation is indicated by the slope of the diagram which could be linear or non-linear. If linear the correlation could be direct or indirect and positive, negative or null. If non-linear, a regression analysis can be used afterwards to measure the strength of the relationship between the variables. The various types of scatter plots are shown in Figure 4.

[FIGURE 4 OMITTED]

The Pareto chart is used to separate the "vital few" from the "trivial many" when analyzing a project. It communicates the 80/20 rule which states that 80 percent of an activity comes from 20 percent of the causes which thus provides rationale for focusing on those vital few activities, Hambleton (2007). The problem frequencies are plotted in the order of greatest to least and show the problems having the most cumulative effect on the system, Kelly and Sutterfield (2006). As mentioned earlier, six-sigma focuses on eliminating as many noise factors as possible and this Pareto analysis helps to do just that. So, the company does not expend resources on wasteful activities that do not add value, but instead cost. to the end project. Figure 5 shows a typical cumulative Pareto chart.

[FIGURE 5 OMITTED]

In the fourth step Improve, the research of the problem's root cause is actually put into work by eliminating all the defects and reducing the degree of variation. Key activities in this phase are performing design of experiments, developing potential solutions, assessing failure modes of potential solutions, validating hypotheses, and correcting/re-evaluating potential solutions. Failure Mode and Effects Analysis is a tool used in this step of the DMAIC process to identify a failure, its mode and effect through analysis. The analysis prioritizes the failures based on severity, occurrence and detection. Through this analysis a company can create an action plan if a failure occurs. Results of an FMEA include a list of effects, causes, potential failure modes, potential critical characteristics, documentation of current controls, requirements for new controls and documentation of the history of improvements. Items listed low on the priority list do not necessitate an action plan to correct them unless they have a high probability of occurrence or are special cause circumstances, Hambleton (2007).

In the fifth and final step Control, the project improvements are monitored to ensure sustainability. Key activities include developing standards and procedures, implementing statistical process control, determining process capability, verifying benefits, costs, and profit growth, and taking corrective action when necessary to bring the project back to its nominal value, iSixSigma (2007). The most important tool used in this phase is the Control chart. It is a statistical method based on continuous monitoring of process variation. The chart is drawn using upper and lower control limits along with a center or average value line. As long as the points plot within the control limits, the process is assumed to be in control. The upper and lower control limits are usually placed three sigma away from the center line because for a normal distribution data points fall within 3-sigma limits 99.7 percent of the time. These control limits are different than the specification limits. Control limits help identify special cause variation and confirm stability while specification limits describe conformance to customer expectations (Skymark Corporation, 2007). However, if there are many outliers, points gravitating toward one of the control limits, or there seems to be a trend in the points, the process may need to be adjusted to reduce the variation and/or restore the process to its center. The control chart can help to assess quantitative gains made through the improvements because each point will be compared to the target value. An example of a control chart for a process in control is shown in Figure 6 below.

[FIGURE 6 OMITTED]

APPLICATION OF METHODOLOGY

Multiple success stories have come from implementing the aforementioned tools into a six-sigma project. The most well known success story is that of General Electric under the direction of Jack Welch. Welch began a six-sigma project to work on the rail car repairs and aircraft engine imports for Canadian customers. Through his guided efforts, the company reduced repair time, redesigned its leasing process, reduced border delays and defects and improved overall customer satisfaction. Quantitatively, General Electric saved over $1 billion in its first year of six-sigma application and then $2 billion in the second year. The company's operating margin rose to 16.7 percent while revenues and earnings increased by 11 and 13 percent respectively, Black, Hug and Revere (2004).

Six-sigma has grown in application from just in the manufacturing environment to also banking, healthcare and automotive. In 2001 Bank of America's CEO Ken Lewis began focusing on increasing the customer base while improving company efficiency. The company handled nearly 200 customer interactions per second and thus set the ultimate goal of its six-sigma efforts to be customer delight. Bank of America established a customer satisfaction goal and created a measurement process to evaluate current performance in order to work toward improving the state. In the first year, the bank's defects across electronic channels fell 88 percent, errors in all customer delivery channels and segments dropped 24 percent, problems taking more than one day to resolve went down 56 percent and new checking accounts had a 174 percent year over year net gain. Within four years of beginning the project, the bank's customer delight rose 25 percent. Through the application of six-sigma, Bank of America was able to focus on the voice of the customer in determining what was most important to them to make their experience a pleasant one, Bossert and Cox (2004).

The Scottsdale Healthcare facility in Arizona began a six-sigma project to work on its overcrowded emergency department because it took 38 percent of the patient's total time within the department to find a bed and transfer the patient out of the waiting room. Before implementing quality efforts, multiple intermediary steps existed in the process which inevitably slowed down the time from start to finish and reduced the potential yield. As a result of the DMAIC and Lean Sigma efforts, the facility identified the root cause of the problem was not that of finding a bed, as originally thought, but rather reducing the number of steps involved in the transfer process. This solution produced incremental profits of $600,000 and reduced the cycle time for bed control by 10 percent. Moreover, the patient throughput in the emergency room increased by 0.1 patients/hour (Lazarus and Stamps, 2002). This project proved one of six-sigma's key arguments that inspection is unproductive and instead quality control should be implemented from the beginning of a product or service to reduce non-value add activities.

It is important to note that not all applications of six-sigma have led to success. A common explanation for the failures is that companies and managers read the latest self-help book and blindly followed the author's recommendations. In doing this, they failed to fully grasp the theory behind the approach. Experience only describes programs like six-sigma, but understanding the theory will help to understand the cause-and-effect relationships which can then be used for rational prediction and management decisions. Also it is important to note that many of the failures are not due to a flaw in six-sigma's conceptual basis, but rather failures in the mechanics of team operation. According to Evans and Lindsay, 60 percent of six-sigma failures are due to the following factors:

* Lack of application of meeting skills

* Improper use of agendas

* Failure to determine meeting roles and responsibilities

* Lack of setting and keeping ground rules

* Lack of appropriate facilitative behaviors

* Ambivalence of senior management

* Lack of broad participation

* Dependence on consensus;

* Too broad/narrow initiatives.

To ensure the success of a six-sigma implementation, it is vitally important that the project champion and other top management officials are very intimately involved in the process. Also implementing constant training for all employees on related topics will minimize the misunderstanding and maximize resources, both financial and human.

CONCLUSION

The objective of this paper was not to examine the tools of Six-sigma in great detail, but rather to explain their history and evolution, because it is more important to understand the origin and development of a discipline, and the reason(s) that it is better than its predecessors. As explained, Six-sigma is used to resolve problems resulting in deviation between what should be happening and what is actually happening. The methodology and tools of Six-sigma can handle any problem type from conformance to efficiency to product/service design. However, in order for the program to effectively mitigate all project risks and be implemented successfully, there must be commitment from the top-down. Through this paper the history of quality has been presented. As its definition and application have evolved from Frederic Taylor's scientific management theories to the present program of six-sigma, and even looking forward to more refined programs such as Fit Sigma, quality continues to be a concept of vital importance for all businesses and industries. In this paper, the underlying theory and concepts contributed by each quality philosopher, and each revolutionary model were explained to show how they are similar but more importantly to show how they are different. For each company and each industry the same quality control methodology may not be appropriate because each operates under different circumstances. The benefits that accrue from applying the different techniques of six-sigma to various production situations are invaluable to a company truly intent upon quality improvement, longevity in its market, and global competitiveness.

REFERENCES

Basu, R. (2001). Six-sigma to Fit Sigma. IIE Solutions, 33(7), 28-34.

Benedetto, A. R. (2003). Adapting manufacturing-based Six-sigma methodology to the service environment of a radiology film library. Journal of Healthcare Management, 48(4), 263.

Black, K., Revere, L., & Hug, A. (2004). Integrating six-sigma and CQI for improving patient care. The TQM Magazine, 16(2), 105.

Bossert, J., & Cox, D. (2005). Driving organic growth at Bank of America. Quality Progress, 38(2), 23-28.

Druley, S. & Rudisill, F. (2004). Which six-sigma metric should I use? Quality Progress, 37(3), 104.

Evans, J.R. & Lindsay, W. M. (2004). An introduction to six-sigma, United States: Thompson Southwestern.

Evans, J.R. & Lindsay, W. M. (2005). The management and control of quality. United States: Thompson Southwestern.

Gitlow, H. S. & Levine, D. M. (2005). Six-sigma for green belts and champions. United States: Pearson Education.

Hambleton, L., (2007). Treasure chest of six-sigma growth methods, tools and best practices. United States: Pearson Education.

iSixSigma (2007). Six sigma DMAIC roadmap. Retrieved January 15, 2008 from http://www.isixsigma.com/library/content/c020617a.asp.

iSixSigma (2002). Sigma level. Retrieved January 15, 2008 from http://www.isixsigma.com/dictionary/Sigma_Level84.htm.

Lazarus, I. R. & Stamps, B. (2002). The promise of six-sigma. Managed Healthcare Executive, 12(1), 27-30.

Samuels, D. I. and Adomitis, F. L. (2003). Six-sigma can meet your revenue-cycle needs. Healthcare Financial Management, 57(11), 70.

Simon, K. (2007). SIPOC diagram. Retrieved January 15, 2008 from http://www.isixsigma.com/library/content/c010429a.asp.

Skymark Corporation (2007). Control Charts. Retrieved January 12, 2008 from http://www.skymark.com/resources/tools/control_charts.htm.

Skymark Corporation (2007). Dr. W. Edwards Deming. Retrieved January 12, 2008 from http://www.skymark.com/resources/leaders/deming.asp.

Skymark Corporation (2007). Philip Crosby: The Fun Uncle of the Quality Revolution. Retrieved January 12, 2008 from http://www.skymark.com/resources/leaders/crosby.asp.

Sutterfield, J.S., and Kelly, C. S. J. (2006). The six-sigma quality evolution and the tools of six-sigma. 2006 IEMS Proceedings, Cocoa Beach, FL, pp. 370-378.

Swinney, Jack (2007). 1.5 sigma process shift explanation. Retrieved January 15, 2008 from http://www.isixsigma.com/library/content/c010701a.asp.

Dominique Drake, Florida A&M University

J. S. Sutterfield, Florida A&M University

Christopher Ngassam, Florida A&M University

Table 1: Defects per Million Opportunities at Various Sigma Levels

Sigma Level Defects per Million Opportunities

One Sigma 690,000
Two Sigma 308,000
Three Sigma 66,800
Four Sigma 6,210
Five Sigma 230
Six-sigma 3.4

Table 2: Defectives per Million Opportunities for Various Process
Off-Centering and Quality Levels

 Quality Level

Off-Centering 3-sigma 3.5-sigma 4-sigma 4.5-sigma

0 2,700 465 63 6.8
0.25-sigma 3,577 666 99 12.8
0.5-sigma 6,440 1,382 236 .32
0.75-sigma 12,288 3,011 665 88.5
1-sigma 22,832 6,433 1,350 233
1.25-sigma 40,111 12,201 3,000 577
1.5-sigma 66,803 22,800 6,200 1,350
1.75-sigma 105,601 40,100 12,200 3,000
2-sigma 158,700 66,800 22,800 6,200

 Quality Level

Off-Centering 5-sigma 5.5-sigma 6-sigma

0 0.57 0.034 0.002
0.25-sigma 1.02 0.1056 0.0063
0.5-sigma 3.4 0.71 0.019
0.75-sigma 11 1.02 0.1
1-sigma 32 3.4 0.39
1.25-sigma 88.5 10.7 1
1.5-sigma 233 32 3.4
1.75-sigma 577 88.4 11
2-sigma 1,300 233 32