In this chapter, we’ll cover the causes and the frequency of defects. And how to create and use a Pareto diagram. This involves listing and categorizing the causes of defects, and then recording the number of defects for each cause category. We then place the causes in descending order, from the category with the highest number of defects to the lowest. Finally, we calculate the total number of defects by all documented causes. In this case, there are 50 defects for Cause 1, 30 for Cause 2, and 10 for Cause 3. We arrive at a sum of 90, this is the total quantity of defects critical for the calculation. From here, we can move on to calculate the percentage of the total that each cause represents. Cause 1 represents 50 defects of the overall 90 defects. If we divide, we find out that 56% of the total population of data, 50 of the 90 of those defects are related to Cause 1. Then we continue the process with all the other causes. For example, with Cause 2 we have 30 defects, that’s about 33% of the total data. And then Cause 3 with 10 defects represents about 11% of the data. We know that we are correct when the percentages from each causes adds up to 100% of the total causes.
The next step is to find the cumulative contributions for each of the causes. What does this mean? Basically, as we add up the percentages of the causes one at a time, starting with cause number one, we’ll get closer and closer to 100% of the defects. Let’s start with the first cause. There was no preceding value, so we only count the percentage of defects for this cause that was 56%. From there we can start to accumulate by adding the 33% for Cause 2. This brings us to 89% accumulated defects for Causes 1 and 2. Finally, we’ll add in 11% for Cause 3 to the 89% we accumulated so far. This gets us to 100%, so now we have the total summation of all the defects that make up our total of 100%.
Next, we can move on to creating the axis and the numeric score scale for our Pareto chart. On the left side, we would have numerical scale quantities. These are the raw number of defects before we calculated their percentages. Then we add bars to represent each category of causes and label them accordingly. From there, we can move on to add the scale on the right axis. The scale on the left axis represented quantities, and on the right axis we have the percentage this quantity represents. So the 50 defects from Cause 1 represents 56% of the total. Then we add Cause 2 with 30 issues and 33% of the total. This brings us to 89% of the total. Finally, we add cause 3, which has 10 defects, representing 11% of the total. This brings us to 100%. As we add the data, the chart becomes very visual. It allows us, at a glance, to determine what the most important causes are. And these are the causes that we should invest our time and effort in fixing.
Pareto’s law, or the 80/20 rule, states that 20% of the causes are very likely to lead to 80% of the defects that we need to address. This 80% are in fact probably the vital few that we need to focus on for continuous improvement. This technique is a very useful quality tool we can use if we have valid and accurate data to help us identify our key defects. In this chart, the red box signifies the defects we should be focusing on.
Now we have another method that’s available to us, it’s called the change-in-slope method. In this method, we examine the cumulative frequency line. And notice the behavior where the slope starts out very steep and then breaks and starts to flatten out. It’s generally those causes on the steep side of the curve where we need to be spending our time and money in improvement projects. This is the best place to start when it comes to process improvement.