The tools I prefer are graphic rather than tabular. The tabular reports give the details if those are needed for further analysis. The graphic reports are intuitive and convey the necessary quality information at a glance. As part of their PMBOK standards, the PMI recommends some basic tools of quality. I've added a tool to the list that I have found particularly useful. With this addition to the current PMI list of tools, the complete list includes the following:
The fishbone diagram (a.k.a. cause-and-effect diagram, a.k.a. spider chart, a.k.a. Ishikawa diagram) is a versatile and intuitive tool. It can graphically present cause-and-effect relationships and root causes. Figure 15-8 is a template of the fishbone diagram. The fishbone diagram can be used when you need to identify, explore, and display the possible causes of a specific problem or condition. It is most commonly used after data compilation and is very effective after an affinity exercise. Teams that are empowered to work on improvement-based projects or are searching for root causes may find this tool helpful. It methodically provides the answer to the general question, “What could be contributing to this problem?” After the general cause is identified, begin asking why. By continuously investigating the why for each cause, the root cause may be more concisely identified.
The fishbone diagram can be used to display the hierarchy of causes of a problem. The true root cause of a problem is not always the obvious option. If you react only to symptoms of a problem, you will need to continually modify your process. Once a root cause is identified (Figure 15-9), resolution of the root cause could provide a permanent fix for a series of annoying problems.
Fishbone diagrams are drawn to clearly illustrate the various causes affecting a process by sorting out and relating the causes. For every effect, there are likely to be several major categories of causes. From this well-defined list of possibilities, the most likely causes are identified and selected for further analysis. When examining each cause, look for things that have changed — deviations from the norm or patterns. Look to cure the cause, not the symptoms.
After a cause-and-effect diagram has been created, the next step is to identify the causes that have the strongest impact on the effect. Often there is existing data that can be used to examine the cause-and-effect relationships, or additional data can be gathered at a reasonable cost and in an acceptable time frame.
If it is not possible to collect data to verify a cause-and-effect relationship, it may be necessary to use process knowledge to select a process variable that is suspected to have a strong influence on a quality characteristic. Once the variable is identified and a way to improve it has been selected, a small pilot test of the process change is conducted, and the results are observed. A successful pilot test provides evidence that there is a cause-and-effect relationship between the process variable and the quality characteristic.
To ensure that you have analyzed a problem fully and correctly, check the proposed root cause against the test criteria listed in Table 15-2. To be a true root cause it must pass all tests. These are listed in Table 15-2.
Control charts track a metric over time. Various patterns that suggest in-control or out-of-control situations are defined. For both situations, some form of follow-up analysis should be done to identify the root causes. The milestone trend chart shown in Figure 15-10 is an example of a control chart.
When you need to identify the actual and ideal path that any product or service follows in order to map process quality and identify deviations and improvement opportunities, the flowchart is another tool that can be beneficial in mapping process quality and performance. It is a picture of steps in a process and can be used to examine the sequence of steps and the relationship between them; to identify redundancy, unnecessary complexity, and inefficiency in a process; and to create a common understanding of the flow of the process.
The flowchart (Figure 15-11) is considered one of the simplest tools. It can be as basic or technically intricate as the process it is used to illustrate. Each type of process step is traditionally identified on the chart by a standardized geometric shape. A flowchart illustrates a process from start to finish and should include every step in between. By studying these charts, you can often uncover loopholes, which are potential sources of trouble. Flowcharts can be applied to anything from the travels of an invoice and the flow of materials to the steps in making a sale or servicing a product.
In process improvement, flowcharts are often used to clarify how a process is being performed or to agree upon how it should be performed. When a process is improved, the changes should be noted on the flowchart in order to standardize the revised flow.
Follow these steps to create a flow chart:
Histograms are probably the most intuitive and flexible of the eight tools presented here. Categories and subcategories for classifying frequency data can be created. Figure 15-12 is a common form of histogram, where the horizontal axis is the data class and the vertical axis is the frequency of observations in each data class. The darker colored bar denotes the average of all the data.
A popular use of the histogram is as an intuitive display of frequency from most frequent to least frequent. This type of histogram is used in Pareto analysis, which is described in the next section.
Pareto analysis is a very simple technique that helps you choose the most effective changes to make.
It uses the Pareto principle — the idea that by doing 20 percent of the work, you can generate 80 percent of the advantage of doing the entire job. Pareto analysis is a formal technique for finding the changes that will provide the biggest benefits. It is useful when many possible courses of action are competing for your attention.
The steps to conducting a Pareto analysis are as follows:
There is always the danger that you will spend a lot of time and money chasing down minor problems while ignoring major problems. Pareto analysis is also referred to as the 80/20 rule. Analysis of data frequently shows that 80 percent of problem occurrences are caused by 20 percent of problem causes. A Pareto analysis produces a histogram showing how many results were generated by each identified cause. If you order this by frequency of occurrence, you can see which problems to attack first. For example, if there were 1,000 failures in a printed circuit, and these are assigned to 15 different causes, which should you attack first? Without some vehicle to help in this selection, it can turn into a duel of personal preferences. If you gather data about how many failures are associated with which cause, it might look like what's shown in Figure 15-13. The graphic display of the data is shown in Figure 15-14.
Milestone trend charts were discussed in detail in Chapter 7. They are quite flexible and can be formatted for a single milestone or multiple milestones. As is true of many of the tools I am offering you, the milestone trend chart is quite intuitive and simple to generate. At a glance, you can see the history of the event you are tracking. The degree to which it is above or below the nominal value and whether it is tracking toward success or failure are also obvious. The run chart (Figure 15-15) can quickly focus your attention on events that need further investigation.
Scatter diagrams originated in the field of statistical analysis. They were used to detect dependent or correlated variables. Underlying the scatter diagram is a mathematical model of the relationship between two variables: a dependent variable, which is represented on the vertical scale, and an independent variable, which is represented on the horizontal scale (see Figure 15-16). Understand that the intention is not to convey a cause-and-effect relationship. This is a common mistake. The scatter diagram merely depicts how the values of two different variables vary from one another. In the example shown in the figure, it would be incorrect to assume that the “total project labor hours” value in any way cause the “# days late” value. The relationship may in fact be causal, but that is not intended nor assumed by the scatter diagram.
Force field analysis is a pictorial representation of countervailing forces that act on an event to either support its occurrence or not support it. Figure 15-17 is the template and Figure 15-18 is an example of its application.
If your intention is to support the occurrence of an event, then you would take actions to support the driving factors and to reduce the restraining factors.
When you need to understand and identify the factors effecting any decision, change, or action, this tool can help build consensus on results found in the fishbone diagram.
A force field analysis is a planning tool for understanding and strategizing about the forces that will drive and restrain any change or implementation plan. It does not illustrate a solution merely through the values. You should use it in the team environment to arrive at areas of consensus based on the team's objectives. It is also a foundation for negotiation.
To conduct a force field analysis, follow these steps:
Use of the force field analysis helps to gain team consensus. In a consensus, the various points of view of each member of the team are considered, discussed, compared, and discussed again until everyone sees all sides of the picture. Ultimately, team consensus leads to a decision that all team members agree on and creates buy-in.
Anytime you implement change, you affect people. By definition, you now have forces supporting and opposing that change. A force field analysis helps identify the forces that resist change as well as those that support the change. This enables you to develop strategies to overcome resistance. Fundamentally, you should do this when implementing any change that involves or affects people.
You should establish trigger values for all of your quantitative metrics. Trigger values serve as early warning signs that indicate the project is heading for an out-of-control situation and some form of corrective action will soon be needed. A trigger value could be a single number such as the frequency of an event in the reporting period, or it could be a trend such as four successive data points moving in the same direction. Figure 15-19 shows an example of a trigger value.
18.216.24.36