Target Stores: The Hunt for “Unvolunteered Truths”

Assignment 1

Target Stores: The Hunt for “Unvolunteered Truths” (10 points)

• Using the Target case, answer questions below (each question is 1 point)
• Download ‘Data_Decision Tree_Target_2017_0912.dta’ and ‘DoFile_Decision Tree_Target_2017_0912.do’ on Canvas
• Assignment format is important to effectively communicate with readers. Badly formatted assignmentswill result in points being deducted from the appearance criterion
• Type, print,and submit your assignment on Wednesday, September20, 2017. A handwritten assignment will be not graded.
• Late assignment will be deducted in the way described on the syllabus

1. What is the main story?

2. Choose two data sources on page 3 and provide four possible variables from each source that Target may collect.

3. How does Target know whether consumers are pregnant?

4. Based on cross-tabulations between ‘pregnant’ and ’12 purchased products’, determine three variables that may be closely related to pregnancy (show the three cross-tabulations and explain your rationale).

5. Explain why ‘Decision Tree analysis’ is applicable to this datasetto predict pregnancy.

6. Using the top three important variables, make a tree graph and explain the classification /splitting criteria associated with each node/segment.

7. Explain the pregnancy rate associated with the characteristics of each segment.

8. Present the classification table (row: actual pregnancy and column: predicted pregnancy) and discuss the usefulness of the model with the correct classification rate and misclassification rate.

9. Think about alife event of consumers (Y) that Target managers may be interested in to increase sales and discuss what products (at least four) may be determinants (X) to predict the event.

10. Replicate Exhibit 3 by hand and explain the role of Guest ID.