Welcome back. Let's talk about factor analysis. Factor analysis is a name given to a class of techniques whose purpose often consist of data reduction and summarization. Ultimately, the goal of factor analysis is to group variables together in order to get to some common meaning. Imagine you asked 200 questions about car buying. The way to use these 200 questions would be to reduce them into two big factors or two big dimensions. To do that, you use factor analysis to summarize the set of observed variables in the form of a smaller number of dimensions called factors. So what is important here is that those factors are latent. Those factors are not directly observed. The two important differences between cluster and factor analysis are that cluster analysis is usually used to group people together whereas factor analysis is used to group variables together to reduce the dimension of the problem or the complexity of the problem. Factor analysis also relies on the identification of latent variables or factors. By the end of this lesson, you'll be able to define factor analysis and provide examples of how it can be used to group variables together. Take a look at this graph. This is an example where factor analysis is used in marketing. What you have here is something called a perceptual map. These maps are obtained by surveying people about various aspects of cars. They can be- there can be question about fuel consumption, affordability, speed, horsepower, luxury, non-luxury, reliability and so on. Then, you want to reduce those questions into two or three main dimensions. When you do the analysis it turns out there seems to be two important factors that would explain how to position brands against one another. One would be price and the other one would be quality. So bear in mind here is that low and high quality could mean something different to different people. Some elements could be objective but some elements could be subjective. There is no one single measure that would define quality for the car. There would be several measures and so you want to summarize these measures into one dimensions. What factor analysis is doing is trying to understand how a customer perceive the quality of the different cars. And the key word here is perceive. Again quality is something very subjective. Similarly, the notion of price or affordability is something that could mean different things to different people. A given price could be very high for someone but very low for someone else. And so, again, to construct these perceptual maps, you need to ask questions to respondents about various dimensions. And then you want to reduce those dimensions of those questions into two important factors or three or four such that you would get as much information as possible. Ultimately, the goal here is to try to understand how I can position different brands of competitors within one simple framework. Here's an example of the results of a factor analysis that was done for shopping experience at retail stores. For example, the important question we ask how would you rank the selection of products in these retail stores? What would be the reputation of these stores? Is it modern not modern and so on. When this analysis was done, the result was such that four factors appear to be important to describe how people value retail stores. And as you can see these four factors are going to be different depending on the variables that inform those factors. So if you look at the first factor that would be for example be the selection of products, quality of products, reputation, and whether or not the store is modern. Here the role of the marketing researcher is then to try to put a name or label to these factor. Similarly factor number two seems to focus on customers, checkout, whether or not it's dull. And the, third one, on cluttered layout and friendliness of clerks. And the last one will be service, location, helpfulness of clerks, and shopping experience. So this last factor, for example, could be called convenience right. The other one could be store environment, for example, and the first one could be product assortment or quality of the product. Overall, this type of analysis can help you summarize several variables into fewer factors. Hence, factor analysis.