So up to now, we have been just setting up the data and setting up the survey results. If you were doing this in real life, you would just put them in a table that looks something like this. In an Excel spreadsheet, import the Excel spreadsheet, and then you could go onto the next function here which is getting the partial utilities. Let's look at that. Here they are. So here are the partial utilities. Known authors is more preferable than unknown authors. Fiction, non-fiction. It seems that non-fiction is more popular. Genre, we have fiction, non-fiction. We did authors, where's the other graph? Oh, it's right here on the bottom right of the screen. So for each of the love attributes, we have a graph like this. This one has three levels. So we can see that it looks like short books are more popular, medium length books are not so popular, and long books are popular here. You can see that utilities are not. That's almost zero right there. We can look at the analysis. These are the average importances, these graphs. Let me run these next few codes and I'll show you what they describe in a little more detail on the slides. There's the importance. Those are my factors and those are my preference plots. Lets obtain the importance of each of these factors, these attributes. So that's what the caImportance does. Let me run the whole line. There we go. Here are my factors and that's the factor importance which will look something like this. Let's look at the slides I have them described in a little more detail. So I put the results in these PowerPoint slides. It's a little more easier to describe. The part utilities here they are. These values represent the part worth of the length of the book, genre, unknown and known author, and then there is this intercept term here. This is the what those graphs we're showing you those part charts. Here we have the actual conjoint analysis results. Notice that it's just a linear model. Here down below you can see the importance of each of these factors. So there they are, they're listed out numerically. Then on the right here, I've listed out all the part-worth utilities. The output shows the estimated impact on utility by each of these factor levels. So if you look at factor X pages one, if it's a short book, it has a negative. Short books not so good, medium length books not so good. Genre 1, I think that was fiction. So in this data set fiction was a positive outcome. Here are the graphs, and this is really where you want to focus your attention. Pages, genre, author, they've have equal weight in this example so consumers primarily look at pages but not by much. Remember, this is a manufactured data set so some of these results are a little squarely. You would imagine that one of these categories would be of more important to people. Here are the part utilities for the level of pages as the results noted. Short medium books are not are not preferred, long books are preferred. Then we can look at a clustering, really what this does is try to group together segments that like similar types of books. Here you can see with two clusters, we can see that there's sort of a black dots and orange dots and those are the two clusters. So there is some grouping. We'll look at that in a minute or in the next video. So how do we interpret the results? We can see that the number of pages in the books has slightly more important than the other two factors, when we look at the rating the attributes, but the most preferred attribute to look is the number of pages. In conclusion, the results indicate that the most popular book would be ones that are written by fiction, known authors, and have more than 1,000 pages.