Six Sigma Statistics with EXCEL and MINITAB

By Issa Bass

Master the Statistical Techniques for Six Sigma Operations, While Boosting Your Excel and Minitab Skills!

Now with the help of this “one-stop” resource, operations and production managers can learn all the powerful statistical techniques for Six Sigma operations, while becoming proficient at Excel and Minitab at the same time.

Six Sigma Statistics with Excel and Minitab offers a complete guide to Six Sigma statistical methods, plus expert coverage of Excel and Minitab, two of today's most popular programs for statistical analysis and data visualization.

Written by a seasoned Six Sigma Master Black Belt, the book explains how to create and interpret dot plots, histograms, and box plots using Minitab…decide on sampling strategies, sample size, and confidence intervals…apply hypothesis tests to compare variance, means, and proportions…conduct a regression and residual analysis…design and analyze an experiment…and much more.

Filled with clear, concise accounts of the theory for each statistical method presented, Six Sigma Statistics with Excel and Minitab features:

• Easy-to-follow explanations of powerful Six Sigma tools
• A wealth of exercises and case studies
• 200 graphical illustrations for Excel and Minitab

Essential for achieving Six Sigma goals in any organization, Six Sigma Statistics with Excel and Minitab is a unique, skills-building toolkit for mastering a wide range of vital statistical techniques, and for capitalizing on the potential of Excel and Minitab.

Six Sigma Statistical with Excel and Minitab offers operations and production managers a complete guide to Six Sigma statistical techniques, together with expert coverage of Excel and Minitab, two of today's most popular programs for statistical analysis and data visualization.

Written by Issa Bass, a Six Sigma Master Black Belt with years of hands-on experience in industry, this on-target resource takes readers through the application of each Six Sigma statistical tool, while presenting a straightforward tutorial for effectively utilizing Excel and Minitab. With the help of this essential reference, managers can:

• Acquire the basic tools for data collection, organization, and description
• Learn the fundamental principles of probability
• Create and interpret dot plots, histograms, and box plots using Minitab
• Decide on sampling strategies, sample size, and confidence intervals
• Apply hypothesis tests to compare variance, means, and proportions
• Stay on top of production processes with statistical process control
• Use process capability analysis to ensure that processes meet customers' expectations
• Employ analysis of variance to make inferences about more than two population means
• Conduct a regression and residual analysis
• Design and analyze an experiment

In addition, Six Sigma Statistics with Excel and Minitab enables you to develop a better understanding of the Taguchi Method…use measurement system analysis to find out if measurement processes are accurate…discover how to test ordinal or nominal data with nonparametric statistics…and apply the full range of basic quality tools.

Filled with step-by-step exercises, graphical illustrations, and screen shots for performing Six Sigma techniques on Excel and Minitab, the book also provides clear, concise explanations of the theory for each of the statistical tools presented.

Authoritative and comprehensive, Six Sigma Statistics with Excel and Minitab is a valuable skills-building resource for mastering all the statistical techniques for Six Sigma operations, while harnessing the power of Excel and Minitab.

• Language: English
• Publisher: McGraw-Hill Education
• Release date: Jul 18, 2007
• ISBN: 9780071542685

Book preview

Chapter 1

Introduction

Learning Objectives:

Clearly understand the definition of Six Sigma

Understand the Six Sigma methodology

Understand the Six Sigma project selection methods

Understand balanced scorecards

Understand how metrics are selected and integrated in scorecards

Understand how metrics are managed and aligned with the organization's strategy

Understand the role of statistics in quality control and Six Sigma

Understand the statistical definition of Six Sigma

A good business performance over a long period of time is never the product of sheer happenstance. It is always the result of a well-crafted and well-implemented strategy. A strategy is a time-bound plan of structured actions aimed at attaining predetermined objectives. Not only should the strategy be clearly geared toward the objectives to be attained, but it should also include the identification of the resources needed and the definition of the processes used to reach the objectives.

Over the last decades, several methodologies have been used to improve on quality and productivity and enhance customer satisfaction. Among the methodologies used for these purposes, Six Sigma has so far proved to be one of the most effective.

1.1 Six Sigma Methodology

Six Sigma is a meticulous, data-driven methodology that aims at generating quasi-perfect production processes that would result in no more than 3.4 defects per 1 million opportunities. By definition, Six Sigma is rooted in statistical analysis because it is data-driven and is a strict approach that drives process improvements through statistical measurements and analyses.

The Six Sigma approach to process improvements is project driven. In other words, areas that show opportunities for improvements are identified and projects are selected to proceed with the necessary improvements. The project executions follow a rigorous pattern called the DMAIC (Define, Measure, Analyze, Improve, and Control). At every step in the DMAIC roadmap, specific tools are used, and most of these tools are statistical.

Even though Six Sigma is a project-driven strategy, the initiation of a Six Sigma deployment does not start with project selections. It starts with the overall understanding of the organization in terms of how it defines itself, in terms of what its objectives are, how it measures itself, what performance metrics are crucial for it to reach its objectives and how those metrics are analyzed.

1.1.1 Define the organization

Defining an organization means putting it precisely in its context; it means defining it in terms of its objectives, in terms of its internal operations and in terms of its relations with its customers and suppliers.

Mission statement

Most companies' operational strategies are based on their mission statements. A mission statement (sometimes called strategic intent) is a short inspirational statement that defines the purpose of the organization and its core values and beliefs. It tells why the organization was created and what it intends to achieve in the future.

Mission statements are in general very broad in perspective and not very precise in scope. They are mirrors as well as rudders: they are mirrors because they reflect what the organization is about, and they are rudders because they point the direction that the organization should be heading. Even though they do not help navigate through obstacles and certainly do not fix precise quarterly or annual objectives, such as the projected increase of Return On Investment (ROI) by a certain percentage for a coming quarter, mission statements should clearly define the company's objective so that management can align its strategy with that objective.

What questions should an organization ask?

Every organization's existence depends on the profits derived from the sales of the goods or services to its customers. So to fulfill its objectives, an organization must elect to produce goods or services for which it has a competitive advantage and it must produce them at the lowest cost possible while still satisfying its customers. The decision on what to produce raises more questions and addresses the nature of the organization's internal processes and its relations with its suppliers, customers, and competitors.

So to define itself, an organization must answer the following questions:

How to be structured?

What to produce?

How to produce its products or services?

Who are its customers?

Who are its suppliers?

Who are its competitors?

Who are its competitors' customers and suppliers?

Figure 1.1

What does an organization produce?

Years ago, a U.S. semiconductor supply company excelled in its operations and was the number one in its field in America. It won the Malcolm Baldrige National Quality Award twice and was a very well-respected company on Wall Street; its stocks were selling at about $85 a share in 1999. At that time, it had narrowed the scope of its operations to mainly manufacturing and supplying electronic components to major companies.

After it won the Baldrige Award for the second time, the euphoria of the suddenly confirmed success led its executives to decide to broaden the scope of its operations and become an end-to-end service provider—to not only supply its customers (who were generally computer manufacturers) with all the electronic components for their products but also to provide the aftermarket services, the repair services, and customer services for the products. The company bought repair centers where the end users would send their damaged products for repair and also bought call centers to handle the customer complaints. About a year after it broadened the scope of its operations, it nearly collapsed. Its stocks plunged to $3 a share (where they still are), it was obliged to sell most of the newly acquired businesses and had to lay off thousands of employees and is still struggling to redefine itself and gain back its lost market share.

At one point, Daimler-Benz, the car manufacturer, decided to expand its operations to become a conglomerate that would include computers and information technology services and aeronautics and related activities. That decision shifted the company's focus from what it does best and it started to lose its efficiency and effectiveness at making and selling competitive cars. Under Jac Nasser, Ford Motor Company went through the same situation when it decided to expand its services and create an end-to-end chain of operations that would range from the designing and manufacturing of the cars to distribution networks to the servicing of the cars at the Ford automobile repair shops. And there again, Ford lost the focus to its purpose, which was just to design, manufacture, and sell competitive cars.

What happened to these companies shows how crucial it is for an organization to not only elect to produce goods or services for which it is well suited, because it has the competence and the capabilities to produce, but it must also have the passion for it and must be in a position to constantly seek and maintain a competitive advantage for those products.

How does the organization produce its goods or services?

One of the essential traits that make an organization unique is its production processes. Even though competing companies usually produce the same products, they seldom use the exact same processes. The production processes determine the quality of the products and the cost of production; therefore, they also determine who the customers are and the degree to which they can be retained.

Who are the organization's customers?

Because an organization grows through an increase in sales, which is determined by the number of customers it has and the volume of their purchases, a clear identification and definition of the customers becomes a crucial part of how an organization defines itself. Not only should an organization know who its customers are but, to retain them and gain their long term loyalty and increase them in numbers and the volume of their purchases, it should strive to know why those customers choose it over its competitors.

Who are the organization's suppliers?

In global competitive markets, the speeds at which the suppliers provide their products or services and the quality of those products and services have become vital to the survival of any organization. Therefore, the selection of the suppliers and the type of relationship established with them is as important as the selection of the employees who run the daily operations because, in a way, the suppliers are nothing but extensions of an organization's operations. A supplier that provides a car manufacturer with its needed transmission boxes or its alternators may be as important to the car manufacturer's operations as its own plant that manufactures its doors. The speed of innovation for a major manufacturer can be affected by the speed at which its suppliers can adapt to new changes.

Most companies have understood that fact and have engaged in long-term relationships founded on a constant exchange of information and technologies for a mutual benefit. For instance, when Toyota Motor Company decided to put out the Prius, its first hybrid car, if its suppliers of batteries had not been able to meet its new requirements and make the necessary changes to their operations to meet Toyota's demands on time, this would have had negative impacts on the projected date of release of the new cars and their cost of production. Toyota understood that fact and engaged in a special relationship with Matsushita Electric's Panasonic EV Energy to get the right batteries for the Prius on time and within specifications. Therefore, the definition of an organization must also include who its suppliers are and the nature of their relationship.

Who are the organization's competitors?

How your competitors perform, their market share, the volume of their sales, and the number of their customers are gauges of your performance. An organization's rank in its field is not necessarily a sign of excellence or poor performance; some companies deliberately choose not be the leaders in the products or services they provide but still align their production strategies with their financial goals and have excellent results.

Yet in a competitive global market, ignoring your competitors and how they strive to capture your customers can be a fatal mistake. Competitors are part of the context in which an organization evolves and they must be taken into account. They can be used for benchmarking purposes.

Who are the competitors' customers and suppliers?

When Carlos Ghosn became the CEO of Nissan, he found the company in total disarray. One of the first projects he initiated was to compare Nissan's cost of acquisition of parts from its suppliers to Renault's cost of acquisition of parts. He found that Nissan was paying 20 percent more than Renault to acquire the same parts. At that time, Nissan was producing about two million cars a year. Imagine the number of parts that are in a car and think about the competitive disadvantage that such a margin could cause for Nissan.

An organization's competitors' suppliers are its potential suppliers. Knowing what they produce, how they produce it, the speed at which they fulfill their orders, and the quality and the prices of their products must be relevant to the organization.

1.1.2 Measure the organization

The overall performance of an organization is generally measured in terms of its financial results. This is because ultimately profit is the life blood of an enterprise. When an organization is being measured at the highest level—as an entity—financial metrics such as the ROI, the net profit, the Return On Assets (ROA), and cash flow are used to monitor and assess performance. Yet, these metrics cannot explain why the organization is performing well or not; they are just an expression of the results, indicators of what is happening. They do not explain the reason why it is happening.

Good or bad financial performance can be the result of non-financial factors such as customer retention, how the resources are managed, how the internal business processes are managed or with how much training the employees are provided. How each one of these factors contributes to the financial results can be measured using specific metrics. Those metrics that called mid-level metrics in this book (to differentiate from the high-level metrics used to measure financial results) are also just indicators of how each one of the factors they measure is performing without explaining why they are doing so. For instance, suppose that the Days' Supply of Inventory (DSI) is a mid-level metric used to monitor how many days worth of inventory are kept in a warehouse. DSI can tell us there is three or four days' worth of inventory in the warehouse but it will not tell us why.

How high or low the mid-level metrics are is also explained by still lower-level factors that contribute to the performance of the factors measured by the mid-level metrics. The lower-level metrics can range from how often employees are late to work to the sizes of the samples taken to measure the quality of the products. They are factors that explain the fluctuations of mid-level metrics such as the Customer Satisfaction Index (CSI). A high or low CSI only indicates that the customers are satisfied or unsatisfied, but it does not tell us why. The CSI level is dependent on still other metrics such as the speed of delivery and the quality of the products. So there is a vertical relationship between the factors that contribute to the financial results of an organization.

A good example of correlation analysis between metrics in a manufacturing or distribution environment would be the study of how all the different areas of operations in those types of industries relate to the volume of held inventory. The higher the volume of held inventory, the more money will be needed for its maintenance. The money needed for its maintenance comes under the form of expenses for the extra direct labor needed to stock, pick, and transfer the products, which requires extra employees; extra equipment such as forklifts, extra batteries, and therefore more electricity and more trainers to train the employees on how to use the equipment; more RF devices, therefore more IT personnel to maintain the computer systems. A high volume of physical inbound or outbound inventory will also require more transactions in the accounting department because not only are the movements of products for production in progress financially tracked but the insurance paid on the stock of inventory is also a proportion of its value and the space the inventory occupies is also rented real estate.

The performance of every one of the areas mentioned above is measured by specific metrics, and as their fluctuations can be explained by the variations in the volume of inventory, it becomes necessary to find ways and means to quantify their correlations to optimize the production processes.

Measuring the organization through balanced scorecards

Metrics are measurements used to assess performance. They are very important for an organization because not only do they show how a given area of an organization performs but also because the area being measured performs according to the kind of metric used to assess its performance. Business units perform according to how they are measured; therefore, selecting the correct metrics is extremely important because it ultimately determines performance.

Figure 1.2 Metrics correlation diagram

Many organizations tabulate the most important metrics used to monitor their performance in scorecards. Scorecards are organized and structured sets of metrics used to translate strategic business objectives into reality. They are report cards that consist of tables containing sets of metrics that measure the performance of every area of an organization. Every measurement is expected to be at a certain level at a given time. Scorecards are used to see if the measured factors are meeting expectations.

In their book, The Balanced Scorecard, Robert S. Kaplan and David P. Norton show how a balance must be instilled in the scorecards to go beyond just monitoring the performance of the financial and nonfinancial measures to effectively determine how the metrics relate to one another and how they drive each other to enhance the overall performance of an enterprise. Balanced scorecards can help determine how to better align business metrics to the organization's long and short-term strategies and how to translate business visions and strategies into actions.

There is not a set standard number of metrics used to monitor performance for an organization. Some scorecards include hundreds of metrics while others concentrate on the few critical ones. Kaplan and Norton's approach to balanced scorecards is focused on four crucial elements of an organization's operations:

Financial

Customer

Internal business processes

Learning and growth

Even though Kaplan and Norton did extensively elaborate on the importance of the suppliers, they did not include them in their balanced scorecards. Suppliers are a crucial element for an organization's performance—their speed of delivery and the quality of their products can have very serious repercussions on an organization's results.

Solectron supplies Dell, IBM, and Hewlett-Packard with computer parts and also repairs their end users' defective products, so in a way, Solectron is an extension to those companies operations. Should it supply them with circuit boards with hidden defects that are only noticeable after extensive use, this can cost the computer manufacturers customers and profit. Motorola supplies Cingular Wireless with mobile phones but Cingular's customers are more likely to blame Cingular than they would blame Motorola for poor reception even when the defects are due to a poor manufacturing of the phones. So suppliers must be integrated into the balanced score cards.

1.1.3 Analyze the organization

If careful attention is not given to how metrics relate to one another in both vertical and horizontal ways, scorecards can end up being nothing but a stack of metrics that may be a good tool to see how the different areas of a business perform but not an effective tool to align those metrics to a business strategy. If the vertical and horizontal contingence between the metrics is not established and made obvious and clear, it would not be right to qualify the scorecards as balanced and some of the metrics contained in them may not be adequate and relevant to the organization.

A distribution center for a cellular phone service provider used Quality Assurance (QA) audit fail rate as a quality metric in its scorecard. They took a sample of the cell phones and accessories and audited them at the end of the production line; if the fail rate was two percent, they would multiply the volume of the products shipped by 0.02 to determine the projected volume of defective phones or accessories sent to their customers. The projected fail rate is used by customer services to plan for the volume of calls that will be received from unhappy customers and allocate the necessary human and financial resources to respond to the customers' complaints.

The projected volume of defective products sent to the customers has never come anywhere close to the volume of customer complaints, but they were still using the two metrics in the same scorecard. It is obvious that there should be a correlation between these two metrics. If there is none, one of the metrics is wrong and should not be used to explain the other.

The purpose of analyzing the organization is primarily to determine if the correct metrics are being used to measure performance and, if they are, to determine how the metrics relate to one another, to quantify that relationship to determine what metrics are performance drivers, and how they can be managed to elevate the organization's performance and improve its results.

What metrics should be used and what is the standard formula to calculate metrics?

Every company has a unique way of selecting the metrics it uses to measure itself, and there is no legal requirement for companies to measure themselves, let alone to use a specific metric. Yet at the highest level of an enterprise, financial metrics are generally used to measure performance; however, the definition of these metrics may not be identical from one company to another. For instance, a company might use the ROA to measure the return it gets from the investments made in the acquisition of its assets. The questions that come to mind would be: What assets? Do the assets include the assets for which the accounting value has been totally depleted but are still being used, or is it just the assets that have an accounting value? What about the revenue—does it include the revenue generated from all the sales or is it just the sales that come from the products generated by a given set of assets?

The components of ROA for one company may be very different from the components of the same metrics in a competing company. So in the Analyze the Organization phase, not only should the interactions between the metrics be assessed but the compositions of the metrics themselves must be studied.

How to analyze the organization

Most companies rely on financial analysts to evaluate their results, determine trends, and make projections. In a Six Sigma environment, the Master Black Belt plays that role. Statistical tools are used to determine what metrics are relevant to the organization and in what area of the organization they are appropriate; those metrics and only those are tracked in the scorecards. Once the metrics are determined, the next step will consist in establishing correlations between the metrics. If the relationships between the relevant measurements are not established, management will end up concentrating on making local improvements that will not necessarily impact the overall performance of the organization.

These correlations between the measurements can be horizontal when they pertain to metrics that are at the same level of operations. For instance, the quality of the product sent to the customers and on-time delivery are both factors that affect the CSI. And a correlation can be found between quality and on-time delivery because poor quality can cause more rework, which can affect the time it takes to complete customer orders.

An example