Hello. Welcome back to this jump tutorial for the design of experiments course. In this tutorial, we're going to talk about fractional factorial design. So we've discussed this factorial design before for this filtration rate experiment. There were four factors; temperature, pressure, mole ratio, and stirring rate. Now we're going to suppose that we only have a half fraction of this design. So we're going to only look at the two to the four minus one design. Using this generator I equals A, B, C, D. So in other words, we're going to confound the main effect D with the A, B, C interaction, and this will become a resolution for design. So we'll have main effects confounded with three-factor interactions, and we'll have two-factor interactions confounded with other two-factor interactions. Here's a table of the design. You can see we have our eight runs because it's a two to the four minus one. We have the basic design in A, B, and C, and then we let D equal ABC. Also, there's a list of the eight treatment combinations that would be included in this design. Of course, I've pulled the data for those eight runs from the original experiment. Jump, I'm going to show you how we can create this design initially, and then of course, how we can analyze it. So let's open up jump. So if we want to create this design from scratch, we can go to the DOE menu, and go down to classical option, and we'll select the screening design. Then once we have this dialogue open, we're going to enter in the number of factors. We have four continuous factors. So you add a four here for the number of factors and continuous. I'm going to change the names to be A, B, C, and D, so they have the names that were used to. I'm going to click continue, and this will give us an option to either choose from a list of fractional factorial designs or construct a main effect screening design. We're going to stick with the first button, choose from a list of fractional factorial designs. So I'll click continue, and here comes our list. So we have different designs for eight runs and 16 runs and then different block sizes. For us, we're going to use this eight run design, and we don't have any blocks in this particular scenario. So we're just going to use this first option here. So I'm going to click continue, and now we can see what our design looks like a little bit. In this display and modify design options here, we can change the generating rules. For this experiment since we have one factor that's confounded, we have that D equals ABC and the positive for that effect. So that's how we're going to generate this design. We can also look at this table of aliasing effects, and see that AB is going to be aliased with CD, AC with BD, and AD with BC. So it's a nice way for us to visualize exactly what this design looks like. Once that looks all good, we can make a randomized run order and make the table. So there's our table with our eight runs and our four factors. I've already created this design and add it in our data. So here's the complete table, and when we want to analyze this, we can go to model, and we have y in as our y role A, B, C, D, and we can look at these three two-factor interactions. So we can click run, and we could look at these effect tests or our parameter estimates. We might notice that A, C, D and AC and AD all look big. That's actually what we found the first time we analyze this with all 16 runs. So that's how we can create a fractional factorial design in jump, and also how we can begin our analysis of one. Thanks.