Pareto Analysis and Pareto Chart
Pareto analysis ranks problems or causes by how much each contributes, so a team can work on the vital few first. A Pareto chart shows the categories as bars in descending order, with a line for the cumulative percentage. It rests on the 80/20 rule: often about 80% of the effect comes from roughly 20% of the causes.
A Pareto Analysis and associated Pareto Chart are essential tools used in lean six sigma to identify comparative frequency of occurrence of a group of measured effects in a process. This technique is used to analyze which effects on a process should be prioritized for improvement. For example, one may want to determine the frequency of observed process failures or the most significant contributing costs of those failures.
Pareto Analysis helps determine whether the frequency of occurrence of failures, the cost of failure, or the effect failures have on downstream processes are of greatest priority to address for improvement. For example, the results of Pareto analysis may show that the most frequently occurring failure is a size variation of a particular part feature, but its relative cost may be insignificant, or its effect on downstream processes may be minimal. Therefore, the analysis should determine that another failure, although occurring less frequently, may have greater potential for cost-saving improvement.
Pareto Analysis is one of the easiest and most effective tools for prioritizing problems. As a result, a Pareto Chart is often used in lean six sigma projects. Pareto Analysis and the Chart should be learned and understood in lean six sigma exam preparation, whether the objective is successful completion of a black belt certification exam or a green belt certification exam.
Pareto Analysis helps determine whether the frequency of occurrence of failures, the cost of failure, or the effect failures have on downstream processes are of greatest priority to address for improvement. For example, the results of Pareto analysis may show that the most frequently occurring failure is a size variation of a particular part feature, but its relative cost may be insignificant, or its effect on downstream processes may be minimal. Therefore, the analysis should determine that another failure, although occurring less frequently, may have greater potential for cost-saving improvement.
Pareto Analysis is one of the easiest and most effective tools for prioritizing problems. As a result, a Pareto Chart is often used in lean six sigma projects. Pareto Analysis and the Chart should be learned and understood in lean six sigma exam preparation, whether the objective is successful completion of a black belt certification exam or a green belt certification exam.
This is the full, step-by-step Pareto analysis with a worked example. For a one-page definition of the chart itself, see what a Pareto chart shows; for the idea behind it, see the Pareto principle (80/20 rule).
Pareto Chart
Let’s suppose we want to know the number of each of three dimensional failures that occur in a process that molds a rubber washer of one inch diameter (outside), ½ inch (inside), and ¼ inch thickness. We take all measurements after molding and vulcanization every ten minutes for eight hours; six per hour. Therefore, we have 48 data points for each feature over eight hours.
A Pareto Chart displays failure modes frequency, usually represented by a bar chart, and arranges the failures in greatest to least occurrence. From the data collected above, and displayed in the chart below, we see that the diameters are usually holding within tolerance for each diameter, but thickness exceeds tolerance consistently, with noted excessive flash at the parting line (not included in the original three search parameters). We would create a simple bar chart of these failures as follows:
Let’s suppose we want to know the number of each of three dimensional failures that occur in a process that molds a rubber washer of one inch diameter (outside), ½ inch (inside), and ¼ inch thickness. We take all measurements after molding and vulcanization every ten minutes for eight hours; six per hour. Therefore, we have 48 data points for each feature over eight hours.
A Pareto Chart displays failure modes frequency, usually represented by a bar chart, and arranges the failures in greatest to least occurrence. From the data collected above, and displayed in the chart below, we see that the diameters are usually holding within tolerance for each diameter, but thickness exceeds tolerance consistently, with noted excessive flash at the parting line (not included in the original three search parameters). We would create a simple bar chart of these failures as follows:
The Pareto Chart indicates that all 48 samples exhibited parting line flash, 29 were oversize in thickness, 4 had oversize o.d. and 1 was undersize i.d. The four failure modes are arranged in order of frequency, greatest to least, for easy prioritization. Analysis may reveal that the last two diameter features were insignificant for immediate improvement consideration. However, the first two, flash, which was not an originally identified failure, and thickness, caused many failures and may be related by cause.
The results of the Pareto Chart and subsequent Analysis in our example determined that the two failures were related. Investigation of other measured and documented data indicated that both failures were caused by lack of sufficient pressure to keep the mold fully closed during injection of molten material. The Pareto Chart was key in prioritizing the several failure modes observed, and consequential Pareto Analysis correctly and quickly identified the cause.
The results of the Pareto Chart and subsequent Analysis in our example determined that the two failures were related. Investigation of other measured and documented data indicated that both failures were caused by lack of sufficient pressure to keep the mold fully closed during injection of molten material. The Pareto Chart was key in prioritizing the several failure modes observed, and consequential Pareto Analysis correctly and quickly identified the cause.
Pareto Analysis and the 80/20 Rule
Pareto Analysis addresses collected data by prioritizing that data according to predetermined factors in a lean six sigma project. If we want to know which failures in a process occur most frequently over a defined period of time, we simply count them as they occur. If we want to know which failures cause the greatest cost to the process, we simply multiply the results of the frequency occurrence with the known cost of each failure. As noted above, some failures are very costly even if they occur infrequently. Therefore, the analysis must be conducted dispassionately. It should be objective and exacting because it makes no sense to attack a problem whose cost of correction in dollars is extreme and saves only a few pennies.
In order to focus on the most important problem factors, the analysis typically employs a simple rule called “Pareto Principle,” also known as the “80/20 Rule.” This rule dictates that 20 percent of causes results in 80 percent of the problems or failures. The implication is that focusing on the improvement of 20 percent (instead of trying to attack all the problem causing factors) will have the best return on the effort and will be the most cost effective.
Both the Pareto Chart and Pareto Analysis are critical tools in lean six sigma projects. In lean six sigma exam preparation, learning about these simple, effective tools will help successfully pass a Lean Six Sigma (LSS) black belt or green belt certification exam.
Pareto Analysis addresses collected data by prioritizing that data according to predetermined factors in a lean six sigma project. If we want to know which failures in a process occur most frequently over a defined period of time, we simply count them as they occur. If we want to know which failures cause the greatest cost to the process, we simply multiply the results of the frequency occurrence with the known cost of each failure. As noted above, some failures are very costly even if they occur infrequently. Therefore, the analysis must be conducted dispassionately. It should be objective and exacting because it makes no sense to attack a problem whose cost of correction in dollars is extreme and saves only a few pennies.
In order to focus on the most important problem factors, the analysis typically employs a simple rule called “Pareto Principle,” also known as the “80/20 Rule.” This rule dictates that 20 percent of causes results in 80 percent of the problems or failures. The implication is that focusing on the improvement of 20 percent (instead of trying to attack all the problem causing factors) will have the best return on the effort and will be the most cost effective.
Both the Pareto Chart and Pareto Analysis are critical tools in lean six sigma projects. In lean six sigma exam preparation, learning about these simple, effective tools will help successfully pass a Lean Six Sigma (LSS) black belt or green belt certification exam.
Additional detail
When you need to deal with different problems at the same time, it might be difficult to know which one you should address first. One of the ways many people do is to use the Pareto analysis. You probably already heard about the 80/20 rule or the Pareto's rule and it is usually used to control the quality. Despite it was originally formulated to study the distribution of wealth, it became known as the Pareto analysis.
The meaning of this rule is pretty straightforward: it simply states that 80% of the quality problems in the final service or product are caused by 20% of the problems in the processes of production. So, as soon as these problems are identified, you'll be much more efficient regarding quality.
The Pareto analysis can be used in many different ways inside your organization. It can help you improve your processes, as already mentioned, it can eliminate the workers who are wasting time by the human resources, and it can be used to discover a big problem that's preventing your company to reach a goal.
Usually, when the Pareto Analysis is done, the results tend to be displayed on a Pareto chart. This kind of visualization of the data has different objectives:
- It arranges the data according to its importance or priority;
- It helps separate the main problems from the smaller ones, so you know where you need to concentrate your efforts;
- It helps you see the real problems and not the perception of a problem since you're basing yourself on real data;
- It allows you to better use your limited resources.
When you use a Pareto chart, you'll be able to see the entire data about the process broken down into the different categories. And since the Pareto chart displays the data in hierarchical order, you'll be able to immediately identify the problems where you need to focus first. Due to the fact that the Pareto chart is very versatile and provides unique information, it's one of the most used charts in management presentations.
According to the Pareto principle, the areas where you have more problems should be the ones that should be addressed first. This way, the Pareto chart is a great way to easily identify those areas.
The next step should be thinking about how to make a Pareto chart.
According to the Pareto principle:
The meaning of this rule is pretty straightforward: it simply states that 80% of the quality problems in the final service or product are caused by 20% of the problems in the processes of production. So, as soon as these problems are identified, you'll be much more efficient regarding quality.
The Pareto analysis can be used in many different ways inside your organization. It can help you improve your processes, as already mentioned, it can eliminate the workers who are wasting time by the human resources, and it can be used to discover a big problem that's preventing your company to reach a goal.
Usually, when the Pareto Analysis is done, the results tend to be displayed on a Pareto chart. This kind of visualization of the data has different objectives:
- It arranges the data according to its importance or priority;
- It helps separate the main problems from the smaller ones, so you know where you need to concentrate your efforts;
- It helps you see the real problems and not the perception of a problem since you're basing yourself on real data;
- It allows you to better use your limited resources.
When you use a Pareto chart, you'll be able to see the entire data about the process broken down into the different categories. And since the Pareto chart displays the data in hierarchical order, you'll be able to immediately identify the problems where you need to focus first. Due to the fact that the Pareto chart is very versatile and provides unique information, it's one of the most used charts in management presentations.
According to the Pareto principle, the areas where you have more problems should be the ones that should be addressed first. This way, the Pareto chart is a great way to easily identify those areas.
The next step should be thinking about how to make a Pareto chart.
According to the Pareto principle:
Stage #1: Identify the possible causes of low quality:
There are many ways to complete this stage. You can do some brainstorming, surveys, focus groups, or apply any other method that you believe will suit the needs. The ultimate goal is to know exactly what the causes may be. Let's say that you manufacture glass and that, somehow, you've been getting many returns because of the cracks. The problem may be on the mishandling when they're still in the factory, the package may not be resistant enough, you can have a poor production process, or there may be problems when they're being transported.
Any of these issues may be the cause or one of the causes of the problem. So, it's time for the next stage.
Any of these issues may be the cause or one of the causes of the problem. So, it's time for the next stage.
When you're looking on how to make a Pareto chart, you need to establish a period on which you are conducting the full analysis and the assessment. What you need here is to have a sample of a determined period of time that will allow you to improve the quality of your product or service. There's only one thing that you need to remember. In case you have a seasonal business, you need to make sure that the data sample that you collect will be appropriate and that will allow you to withdraw the right conclusions.
Once you selected the time period, you will need to make the necessary match to one of the causes of the low quality that you identify during the first stage. If you look at the previous example, when you get a return stating that there was a misleading in the factory, you need to add one to this cause. If you get a returned glass that says the package wasn't strong enough, then add one here as well.
This stage only ends when the period of time that you pre-determined to get your sample data is finished.
Once you selected the time period, you will need to make the necessary match to one of the causes of the low quality that you identify during the first stage. If you look at the previous example, when you get a return stating that there was a misleading in the factory, you need to add one to this cause. If you get a returned glass that says the package wasn't strong enough, then add one here as well.
This stage only ends when the period of time that you pre-determined to get your sample data is finished.
During this stage, you'll need to get the subtotals for each different cause of the poor quality of your product. And when you have all these subtotals, its sum will give you the total number of defects that your product had in that specific period of time.
But how can you make a Pareto chart from here? The first thing you'll need to do is to acquire the percentages of each cause, in relation to the total number of defects. All you have to do is to divide each subtotal by the total number of defects and multiply by 100.
Once you get all the percentages, you can finally draw your Pareto chart in a histogram. You'll use the causes of the poor quality on the x-axis, and on the y-axis, you'll have the percentages of their occurrences.
The most often cause should be put on the left, and the others should follow in that order. To make sure that you can immediately see what's happening, you should also draw a cumulative line graph that should be placed over the histogram bars.
But how can you make a Pareto chart from here? The first thing you'll need to do is to acquire the percentages of each cause, in relation to the total number of defects. All you have to do is to divide each subtotal by the total number of defects and multiply by 100.
Once you get all the percentages, you can finally draw your Pareto chart in a histogram. You'll use the causes of the poor quality on the x-axis, and on the y-axis, you'll have the percentages of their occurrences.
The most often cause should be put on the left, and the others should follow in that order. To make sure that you can immediately see what's happening, you should also draw a cumulative line graph that should be placed over the histogram bars.
Stage #4: Interpretation and applying the results:
In this stage, it's important to have present the 80/20 rule. You'll notice that, in most cases, 80% of the products with defect were caused by around 20% of the possible causes.
One of the things that you should have in mind is that after you have implemented the solutions for the first cause of defects, you should do another Pareto Analysis. The percentages relative to the other causes will be substantially different so, the effort is worthwhile.
The Pareto chart isn't the only diagram that can be used to look for the problems in your processes that can affect your products or services quality. Learn about various diagrams used in DMAIC six sigma methodology.
One of the things that you should have in mind is that after you have implemented the solutions for the first cause of defects, you should do another Pareto Analysis. The percentages relative to the other causes will be substantially different so, the effort is worthwhile.
The Pareto chart isn't the only diagram that can be used to look for the problems in your processes that can affect your products or services quality. Learn about various diagrams used in DMAIC six sigma methodology.
Related guides
Last updated: · Published by Lean Six Sigma Training Guide (sixsigmatrainingfree.com) · How this guide is written