1

Instructions for Questions 1 - 6

Georgia Tech Bank (GT Bank) just hired you as a consultant, to help them better understand their business. On your first day, you decide to use the "Credit" dataset from the ISLR package, as a proxy, to answer questions about their customer base. Use the "Credit" dataset from the ISLR package, drop the ID column, and model the data (linear-linear model) such that Balance is your dependent variable and the rest of the features (Income, Limit, Rating, Cards, Age, Education, Gender, Student, Married, and Ethnicity) are your independent variables. Use your model to answer questions Q1 through Q6.

Data Dictionary for the "Credit" Dataset:

1. Balance: This is the average credit card balance in dollars for each individual. 2. Income: The annual income of the individual in thousands of dollars. 3. Limit: The credit limit on the individual's credit card. 4. Rating: The individual's credit rating. 5. Cards: The number of credit cards owned by the individual. 6. Age: The age of the individual. 7. Education: The highest level of education completed by the individual. 8. Gender: The gender of the individual (e.g., "Male" or "Female"). 9. Student: Indicates whether the individual is a student (e.g., "Yes" or "No"). 10. Married: Indicates whether the individual is married (e.g., "Yes" or "No"). 11. Ethnicity: The ethnicity of the individual, which can include categories such as "African American," "Asian," and "Caucasian."

Note: To access it, simply install the ISLR package in R (if you haven't already) and load the dataset using the following commands: install.packages("ISLR") library(ISLR) data(Credit)

Credit

1 point

Question 1 is unpinned. Click to pin.

Question at position 1

Which features are significant at an alpha level of 0.001 ?

Which features are significant at an alpha level of 0.001 ?

MarriedYes, StudentYes, Cards

Rating, Limit, EthnicityAsiam

Income, Age, Rating, Education

Income, Limit, Cards, StudentYes

Question at position 2

2

1 point

Question 2 is unpinned. Click to pin.

Question at position 2

What is the absolute difference in coefficients between a customer of Asian ethnicity and Caucasian ethnicity? Round your answer to two decimal places.

What is the absolute difference in coefficients between a customer of Asian ethnicity and Caucasian ethnicity? Round your answer to two decimal places.

26.90

16.80

10.11

6.70

Question at position 3

3

1 point

Question 3 is unpinned. Click to pin.

Question at position 3

Which statement about Rating is correct?

Which statement about Rating is correct?

Rating is significant at a 0.01 level of alpha

For one unit increase in rating, the average balance decreases by 1.13653 given all other features are held constant.

The base case for rating is -479.20787

None of the above

Question at position 4

4

1 point

Question 4 is unpinned. Click to pin.

Question at position 4

Which features have a high level of multicollinearity, with a VIF over 5?

Which features have a high level of multicollinearity, with a VIF over 5?

Limit, Cards, Age

Gender, Education

Limit, Rating

Age, Cards

Question at position 5

5

1 point

Question 5 is unpinned. Click to pin.

Question at position 5

How many influential points are included in the dataset, with a Cook's distance over 1?

How many influential points are included in the dataset, with a Cook's distance over 1?

0

1

3

4

Question at position 6

6

1 point

Question 6 is unpinned. Click to pin.

Question at position 6

Which statement correctly describes the residual vs fitted plot for the model?

Which statement correctly describes the residual vs fitted plot for the model?

The residuals are randomly scattered around 0, suggesting that the model is a good fit.

The residuals show a distinct "V" pattern, suggesting that model has room for improvements.

The residuals are as high as 300, suggesting that the model is suffering from variance.

The residuals are as low as -150, suggesting that the model is suffering from bias.