Description
Group Assignment II
All Data files are on the class web page.
1. Mass Transit: Through a survey of approx. 600 consumers, the Mass Transit Authority has collected information on consumers’ preference for use of Public Transportation as a function of gasoline prices. In addition the Mass Transit Authority collected attitudinal/demographic data on these same consumers.
There are two data files: (a) calibration.xls that has customer id their preference for using public transportation (x1-x12) along with their demographic characteristics (b) prospect.xls has customer id and demographic information pertaining to a list of prospective users. See data description file for additional details.
The department is seeking your advice on whom to target. Base your decision on the following analysis:
(a) Use the primary data set on usage intention (calibration.xls) to perform cluster analysis. Specifically, use variables x1-x12 to perform this analysis and segment customers into two groups – potential users and non-users of public transportation.
(b) Using the demographic variables in calibration.xls perform Discriminant analysis to identify the variables that best discriminate potential users from non-users. Use the Discriminant Function from the above to classify consumers in the prospect.xls into the appropriate groups (testdata option can help you do that in one shot).
(c) Based on the results in (b) the Mass Transit Authority engaged in a campaign to promote the use of their mode of transportation amongst the 300 consumers in the prospect list. They tracked the behavior of these 300 consumers and created a variable used = 2, if the consumer used mass transit and is set to 1 otherwise. This information is available in validation.xls. Compare your predictions from (b) with the actual behavior of the prospects to see how well you did in classifying users and non-users.
Interpretation of the Results
(i) Based on your results in (a) how would you determine which cluster represents consumers that are users (non-users) of public transportation?
(ii) What criteria would you use to assess the goodness of your segmentation in (a)?
(iii) Based on your results in (b) which demographic variables help discriminate potential users of public transportation from non-users. Why?
(iv) How good is your discriminant analysis classification – what criteria would you use to ascertain this?
(v) Comment on the cross-tabulation in (c). What does that tell you about the effectiveness of your classification procedure.




Reviews
There are no reviews yet.