Some of the most popular techniques for identifying the causes of quality problems are Pareto charts, flowcharts, check sheets, histograms, scatter diagrams, process control charts, and cause-and-effect diagrams. These well-known tools are sometimes known as the "magnificent seven," the "seven QC tools" and the seven process improvement tools. We discuss each in the following sections, and they are summarized in Animated Figure 3.4.
Pareto analysis is a method of identifying the causes of poor quality. It was devised in the early 1950s by the quality expert Joseph Juran. He named this method after a nineteenth century Italian economist, Vilfredo Pareto, who determined that a small percentage of the people accounted for most of the wealth. Pareto analysis is based on Juran's finding that most quality problems and costs result from only a few causes. For example, he discovered in a textile mill that almost 75 percent of all defective cloth was caused by only a few weavers, and in a paper mill he studied, more than 60 percent of the cost of poor quality was attributable to a single category of defects. Correcting the few major causes of most of the quality problems will result in the greatest cost impact.
Pareto analysis can be applied by tallying the number of defects for each of the different possible causes of poor quality in a product or service and then developing a fre-quency distribution from the data. This frequency distribution, referred to as a Pareto diagram, is a useful visual aid for focusing on major quality problems.
Consider a product for which the causes of poor quality have been identified as follows.
For each cause of poor quality, the number of defects attributed to that cause has been tallied over a period of time. This information is then converted into the Pareto chart shown in Figure 3.5.
This Pareto chart identifies the major cause of poor quality to be poor design. Correcting the design problem will result in the greatest quality cost savings with the least expenditure. However, the other problems should not be ignored. TQM teaches us that total and continual quality improvement is the long-term goal. The Pareto diagram simply identifies the quality problems that will result in the greatest immediate impact in quality improvement.
A flowchart or process flowchart is a diagram of the steps in a job, operation, or process. It enables everyone involved in identifying and solving quality problems to have a clear picture of how a specific operation works and a common frame of reference. It also enables a process improvement team to understand the interrelationship of the departments and functions that constitute a process. This helps focus on where problems might occur and if the process itself needs fixing. Development of the flowchart can help identify quality problems by helping the problem solvers better understand the process. Flowcharts are described in greater detail in Chapter 6 ("Process Planning and Technology Decisions") and Chapter 8 ("Human Resources in Operations Management").
Check sheets are frequently used in conjunction with histograms, as well as with Pareto diagrams. A check sheet is a fact-finding tool used to collect data about quality problems. A typical check sheet for quality defects tallies the number of defects for a variety of previously identified problem causes. When the check sheet is completed, the total tally of defects for each cause can be used to create a histogram or a Pareto chart.
Scatter diagrams graphically show the relationship between two variables, such as the brittleness of a piece of material and the temperature at which it is baked. One temperature reading should result in a specific degree of brittleness representing one point on the diagram. Many such points on the diagram visually show a pattern between the two variables and a relationship or lack of one. This diagram could be used to identify a particular quality problem associated with the baking process.
We discuss control charts and other statistical quality control methods in Chapter 4, "Statistical Quality Control." For now, it is sufficient to say that a control chart includes a horizontal line through the middle of a chart representing the process average or norm. It also has a line below this center line representing a lower control limit and a line above it for the upper control limit. Samples from the process are taken over time and measured according to some attribute. In its simplest form, if the measurement is within the two control limits, the process is said to be in control and there is no quality problem, but if the measurement is outside the limits, then a problem probably exists and should be investigated.
Statistical quality control methods such as the process control chart are important tools for quality improvement. Japanese employees at all levels, and especially in production, are provided with extensive training in statistical quality control methods. This enables them to identify quality problems and their causes and to make suggestions for improvement.
A cause-and-effect diagram, also called a "fishbone" diagram, is a graphical description of the elements of a specific quality problem and the relationship between those elements. It is used to identify the causes of a quality problem so it can be corrected. Cause-and-effect diagrams are usually developed as part of participative problem solving to help a team of employees, supervisors, and managers identify causes of quality problems. This tool is a normal part of the problem-solving activity of quality circles in Japanese companies; however, in the United States it is often used separately, outside the quality circle program.
Figure 3.6 illustrates the general structure of the cause-and-effect diagram. The "effect" box at the end of the diagram is the quality problem that needs correction. A center line connects the effect box to the major categories of possible problem causes, displayed as branches off of the center line. The box at the end of each branch (or fishbone) describes the cause category. The diagram starts out in this form with only the major categories at the end of each branch. Individual causes associated with each category are attached as separate lines along the length of the branch during the brainstorming process. Sometimes the causes are rank-ordered along the branches in order to identify those that are most likely to affect the problem. The cause-and-effect diagram is a means for thinking through a problem and recording the possible causes in an organized and easily interpretable manner. The causes listed along the branches under each category in Figure 3.6 are generic, but for a specific quality problem with an actual product, the causes would be product and problem specific.
*3-30. For a look at how the seven tools are applied in the software industry, link to this case study about advanced RISC machines.
*These exercises require a direct link to a specific Web site. Click Internet Exercises for the list of internet links for these exercises.