Data-based testing of ideas and improvements. This is one of the principles of Lean Six Sigma. Also this principle is sometimes referred to as evidence-based improvement actions. But what does data-based testing of ideas mean? How should you do this in practice? Well, in this introduction video on data analysis, I will show you what data-based testing is. First, you will learn how to model your CTQ and your influence factors. Next, you will learn to identify several tools for data-based testing, and finally, I will show you when these tools are applicable. This is an introduction video in which I will show you the building blocks and the foundations of database testing. Usually, you start analyzing data if you want to know if two different concepts are related to each other. In a Lean Six Sigma project, you are wondering if your CTQ is related to an influence factor. If you look at the DMAIC phases of your project, this question will be part of the improve phase where you want to establish the effect of an influence factor on your CTQ in order to develop and implement an effective improvement actions. Fourth, data based testing of ideas and improvements. You are interested to find data that show that your influence factor. X has an effect on your CTQ, your Y in your project. Using data, you will test if this relationship exists or not. Let's illustrate this with some examples. Remember that we looked at the caffeine percentage in coffee beans. Now we would like to know which factors determine the caffeine percentage level in the beans. The caffeine percentage is therefore your CTQ or your Y variable. And the factors that can possibly affect the degree of caffeine in the beans are the X variables. For example, the number of extractions or the duration of extraction process or the amount of solvent that is used during the extraction process. As another example, consider patients in a hospital, you would like to know which factors determine the length of their stay to see if you can improve this. Lengths of stay is therefore the CTQ or your Y variable. Possible X variables, also called influence vectors, are the type of appointment schedule that is used. The age of the patient or the type of surgery they had. To analyze. If two variables are related, you have to determine which variable is your influence factor or your X variable and which variable is your CTQ or your Y variable. Next, you can model the relationship using data. So, now that we know that you have to model your CTQ and your influence factor, the next step is to learn which tool can we use and when should you use these tools. There are many methods to analyze data. I will show you regression, ANOVA, logistic regression, and chi-square test. The suitable method depends on the type of data that you have. It depends if your CTQ and your influence factor are categorical or numerical variables. Let's take a look when to use which method. If the Y variable is numerical and your X variable is also numerical, you should perform a regression analysis. This analysis will be explained in these videos. If your Y variable is numerical and your X variable is categorical, you should perform an ANOVA, which will be explained in these videos. A special case of the ANOVA is a 2-sample t-test. If the assumptions underlying the ANOVA are not met, you can perform a Kruskal-Wallis analysis which will be explained in this video. If the Y variable is categorical and your X variable is numerical, you can perform a logistic regression. In this video that will be explained. Finally, if the Y variable is categorical and your X variable is also categorical, you can perform a chi-square analysis. Which will be explained in this video. Summarizing. To analyze if two variables are related, you have to determine which variable is your influence factor, or your X variable, and which variable is your CTQ, or your Y. Next, you have to determine whether your variables are categorical or numerical. Once you have determined what type of data you have, the tree diagram can be a very useful tool to determine which type of analysis method is suitable for your data. And these are the methods that you can use for that data-based testing of ideas and improvements in your Lean Six Sigma project.
