a.
a.You work for an airline and the customer database has information on Relationship Status as well as amount paid for recent travel tickets. Relationship status is coded as: 0:Single, 1:Married, 2: Divorced, 3:Widowed, 4:Commited, 5:It's Complicated. Amount spent is in the database as actual dollars paid. Your Chief Marketing Officer wants to determine whether there is a correlation between amount spent and the customer's relationship status. She asks you to calculate the correlation between relationship-status and amount spent. Is this a good idea, or not? Explain. b.You are in a meeting with the CMO of a major retailer discussing a simple regression model to predict Amount_Spent (Y) from Age (X). The model has a high r-square value, and everything about the final model is statistically significant. The coefficient for AGE is 0.00000001234. The CMO points out that the coefficient for AGE is almost zero, and anything multiplied by zero is zero; So, AGE really doesn't matter in the model. Agree? Disagree? How would you respond?
a) Its a good idea. To know whether there exists such a correlation or not can be considered valuable information with the help of which potential valuable customers can be found and leveraged. For example, if we find a correlation that married/committed customers spend more, we can specially target these customers with offers etc and "capture" their business.
b) Context matters. Age is a variable not crossing more than 100 in most cases. So the amount in dollars (which we assume Y is, this is also important) is not affected much by 0.00000001234*AGE. Therefore, I agree with the CMO, AGE factor is not really that important. I agree with him.
Hope this was helpful. Please do leave a positive rating if you liked this answer. Thanks and have a good day!
Get Answers For Free
Most questions answered within 1 hours.