In this session, you're going to look at the first part of root cause analysis. The first part of root cause analysis is trying to formulate hypothesis. What are the things that we should be testing in terms of being the causes of the effects that we're seeing in terms of something that a process customer cares about? How do we go about coming up with the idea that something might be affecting the output that we care about? That's the idea of formulating hypotheses. That's what happens in the analyze phase of a Six Sigma project. In the analyze phase of a Six Sigma project, you try to get information about what may be some of the things that we may want to change in the way we do things so that we get better results for the process customer. If you think about Six Sigma projects, they are made up of cross-functional teams. You're trying to get people from different functions, from different levels of expertise, from different areas of expertise together. We're trying to get all of their knowledge to come together and come up with informed hypotheses on what might be things that might be related to each other and that they should be looking at. Then you want to take all of that information and next go into testing those root cause analysis and cause-effect relationships in the form of statistical tests, hypotheses tests, or even doing experiments and seeing if something is what you were expecting it to be based on your analysis. In this session, we're going to look at some techniques that are used in order to gather that data from the experts. Gather that data from a bunch of people who are coming from different areas, who have different perspectives. What is the main idea? The main idea is that there is a Y value or a Y, which is the effect that is caused by certain causes. Those might be the different Xs and they might be many Xs and there might be interactions among those Xs and that's what we want to try and parse out. We want to be able to say, how is the Y affected by the different Xs. If you think about Six Sigma projects, if you think about process improvement projects, you can think about it from the point of view of here's the Y, here's the main X that is causing that Y, and then you could start breaking it down. You can have cascading sets of Y as a function of X equations where the Y is a function of X as the top equation. Those Xs becomes the Y for the next level, so you're basically doing some sort of a 5-Why Analysis in terms of drilling down to the Xs that you want to control so that you affect the Y. If you think about Y and X, you can be thinking about this Y as the dependent variable, X as the independent variable. Y as the outcome, X as what you're doing in terms of trying to create that outcome. Y as a symptom of a problem. Y is the number of defects that we see, and you could be looking at the causes in terms of what are the problems in the process itself that are causing those defects. Finally, you could be thinking about in terms of, Y is something that you monitor. It's already the effect. It's too late when you get that Y to make a change to it. It should be the Xs that you should be controlling and the Ys that you should be monitoring. You should be constantly testing the relationship between the Xs and Ys in order to have your process working at a certain level. But that's the main idea of root cause analysis, Y is a function of many Xs. Some standard things to think about when you are thinking about different types of Y. Here you see a set of Ys and the corresponding big Xs that you may be considering. Just to get some sense of what are we talking about as Ys and Xs here, Y could be customer satisfaction that is affected by out-of-stock items. When you have a lot of out-of-stock items, your customers are going to be dissatisfied. Y could be the expense in terms of how much you have to spend on stuff that is waiting, stuff that is being kept in storage, and the X could be the amount of work in process which stands for the work-in-process inventory. Production cycle time could be the Y that you care about and that will be affected by the amount of rework. Because as you have to do something over and over again, if you don't do it right the first time, it affects your production cycle time. It increases your production cycle time because you're using up your capacity for other things. That's an X that you try to reduce in order to get a better level of Y. Defect rate as the Y being affected by inspection procedures. If you're thinking about lean as a philosophy, there would be saying that, hey, you shouldn't be thinking about inspection procedures in the first place. But, if you do have to have inspection because you don't have perfect processes yet, or it's a new process, and you still have inspection procedures, that's something that you might want to focus on in order to have less defects or a smaller defect rate. Process temperature affecting the dimension of a part. These could be environmental things that you may want to control. Right now you're just saying that there might be an effect of process temperature on dimension of part, you figured out if there is, then that's something in the environment that you need to control in order to get better quality outputs for your customers.
