QuikSigma: Pareto Charts for Six Sigma Projects

Pareto charts for Six Sigma projects are great tools to help you isolate problems within your company. In the video below, we go over how to use Pareto Charts to find where, when, and by whom defects are created. QuikSigma features patented multivariate Pareto Charts which greatly expand the insight this tool provides.




In the measure phase of the project, we’re going to expand the baseline statement that we made in the charter. So we’re going to explain more about it, talk about whether it’s stable and predictable, and where and when it happens and those kinds of things. The Pareto chart is one of the tools that’s available for that, and it shows you where and when and by whom problems are created.

So I’ve arranged some data down here for us, some tabs that we can look at. In the simple form, what you’re going to do is see something like this where we’ve got count. That’s obviously our Y variable, whoops I think I missed and this can be either an XC variable or it can be a block and I’ll just show you the difference. If I make that an XC variable in the simple case, like that, I get my sorted histogram. So my largest occurrence first and then working down and then I get a cumulative sum line on the top. On the other hand, if I make that a block and hit calculate, I get the same sort of thing. I get my sorted histogram, but I get rid of the cumulative sum. So that’s the simple version. Let’s look at a little more complex version. Let’s say that we’re worried about bubbles. So that’s my Y variable and I’ve got my choice of three different input variables. So if I want to look, for example by machine, I can make machine the block variable and lo and behold, I get the information that most of my bubbles are happening on the Corona machine and that the Smith machine is running the least and that can cause me to go ask some very high-quality questions. That’s one reason that a Pareto chart is so widely used in both Lean and in Six Sigma. Now, here’s a subtlety, suppose that I want to look at combinations of variables. So I will leave this, the machine type, the block, and I’m going to make material an XC variable. Now watch what happens. The outer beige columns are my old blue columns when I ran just the Machine variable but nested within them, I have my material type. So I can look and see, well yeah, I really get into problems when I run polycarbonate on the Corona machine, but look, the Royal machine doesn’t seem to have any problem with polycarbonate. Its problems are elsewhere.

Now, one last thing. If I designate two or more XC variables, so now I’ve made both material and machine and XC. Ok, now across the x-axis it will put all combinations of the input variables. So I can see that my number-one problem is the Corona polycarbonate. Probably more often though you’ll make one a block and one an XC and look at combinations. That way that’s just really a very useful display.

Return to Blog Posts >>

Leave a Reply

Your email address will not be published. Required fields are marked *