Compare and Contrast the Input Data for Cluster Analysis, Choice Models, and Positioning Analysis

1. Compare and contrast the input data for cluster analysis, choice models, and positioning analysis.

Cluster analysis is taking segmentation data based on observations from a number of respondents to put them into similar clusters in order for us to identify segments. The data used in choice models is based on just that, "choice". Based on customer choices we are able to group those consumers based on their previous choices such as past purchase history, annual spending, and where they are located. Now for positioning analysis we have perceptual data similar to the data used in the segmentation analysis but this is data that shows us how consumers perceive the various brands being compared based on certain attributes. With positioning analysis we also have preference data on consumers that shows us for each respondent how they rank the brands by which offering (brand in this case) they prefer. At times this data can look similar but they are particularly different and lead you to very different results based on the analysis you are running. 

1. Should you use a two-dimensional or a three-dimensional map?  Please justify your choice.

1. Here are some pointers.:

• Produce 2-D and 3-D maps.  Compare the diagnostics.

• Also think about the differences in the strategic implications of the two maps. Which dimension was not picked up in the 2-D map?  Is that important?  For instance, if price is an important factor for consumers and is not being picked up in the 2-D map, then we should use the 3-D map.

Looking at the diagnostics of both outputs we only see a difference in the proportion variance explained. In the 2D map we have a 0.974 proportion variance explained and 0.987. Because the variances are so close this is an indication that we would want to use the 2D map for the sake of simplicity. If we are not gaining much more if any insight using the 3D map, it is not necessary and will only add complexity that is not needed. The dimension shown on the 3D map that isn't on the 2D is showing the attributes too close together to easily discern what is of importance on that 3rd map but from what I can tell is that it is showing price but I believe this is something we can gather some insight from in the 2D map. It does not provide much additional insight that will be necessary so with that being said I think it is safe to say using a 3D map is not worth the extra effort and a 2D map gives us everything we need for the positioning analysis.

1. Run the analysis using either 2-D or 3-D using the vector preference model.  Use the information in case and the model output to answer the following questions: Name the two (or, three) dimensions underlying the perceptual maps that you generated.

We are using a 2D map, so we have two dimensions that we can name. Based on the layout of the attributes we can see that overall, quiet, prestige, economical, common, interesting, successful, poorly built, and uncomfortable fall on the X-axis. The dimension here can be called the "customer brand quality" dimension. These dimensions show us how respondents perceive the brands in terms of quality. 

The Y-axis has attributes such as roomy, avantgarde, attractive, sporty, easy serviceable, and poor value. These indicate that the y-axis deals more with the look of the vehicles offered. This dimension can be called the "Design" dimension. The X-axis has a 54.6% variance while the Y-axis has 18.8% variance. This shows that the X-axis holds more than double importance in explaining the respondents perceptions