80% of all problems can be attributed to 20% of the causes. Ever heard this saying before?, the 80/20 principle, also known as the Pareto principle, and also known as the rule of the vital few and the trivial many? It is this rule of thumb that can be very useful in solving problems. And therefore, I want to explain to you what a Pareto chart is, a chart in which a Pareto principle can easily be spotted. Of course, I will also show you how to make such a chart using Minitab. Consider, you work in a factory that produces cookies called Cafe Noir. These are famous Dutch cookies and they are really, really good. Sometimes the production line is stopped due to some problem, and you have recorded data on these production stops in a data file, which looks like this. For each stop, you have recorded the duration of the stop in minutes as well as the cause for the stop. You start a project with the objective to reduce the time lost due to production stops. Therefore, you wish to select the cost that occurs most often. Using a Pareto chart, we can analyze this. Now pause this video, open your Minitab, load your data before you continue. The data made up should look like this with Stops in the first column, Duration in the second column, and Cause in the third column. Note that the third duration is missing, which is indicated with a star. If you scroll down, you will see that we have in total measured 90 stops. Okay, to make a Pareto chart of the causes, you go to the Start > Quality Tools > Pareto Chart. Now Minitab asks for you, what is your defects or attribute data, which is basically the category that you want to count. That is, causes. Okay, we now get the Pareto chart which counts how often each cause on the x-axis occurs. Before we discuss this Pareto chart in detail, I want to make a second Pareto chart which weighs the causes with their durations. For that we go back to our menu, which is most easily done with Edit Last Dialogue button. Now instead of frequencies, which we left empty, we now put Durations. So we will weigh the causes with a duration. Okay, this is our second Pareto chart, which has now duration on our horizontal axis. But let's first study the count chart. This is the chart we made. Do you already see what the main issues are? Let's have a closer look at the different parts in the chart. The horizontal axis contains the type of problems. The height of the bars in the chart indicates how often these problems occur. The bars are sorted in decreasing frequencies, such that the most frequently occurring problems are placed to the left, and the many minor issues are placed to the right. Note that the smallest issues are combined into a single category that is labeled Other. The frequencies can be read from the vertical axis on the left, as well as from the count on the bottom of the chart. See that the problem deviation and dimensions occurs 37 times, and remember that we have included 90 production stops, so this problem is a large cause. The curve shows the cumulative frequencies of the issues. These cumulative percentages can be read from the right vertical axis as well as from the bottom of the chart. From this chart, it can be concluded that the three largest issues cause 80% of all problems. However, this Pareto chart with the frequencies only shows how often each problem occurs. But most of the time we are more interested in which problem takes up most of our time. We also measure the duration and main operator chart with this variable. Let's take a look at this output. This Pareto chart looks similar to the one with the frequencies. The bottom row now shows the total times several problems took instead of the total number of times they occurred. In the bottom row, we can see that the breaking of the cookies might occur less often than deviations, but takes up most of the time. We can also see that the four biggest problems cause 79.2% of the time that is lost due to production stops. This is the 80/20 principle. Only a few issues cause the majority of the problems. However, it will never be exactly 80 and 20%. Of course, it could be that, say, 70% of your problems can be attributed to only 10% of the causes. The main principle is, there are always a few vital issues and trivial many, and make sure to focus on the vital few. Summarizing, a Pareto chart can help to distinguish the vital few causes from the trivial many. We can do this, either by counting the number of occurrences, or by weighing them with the relevant characteristic.
