You work for a regional bank, which has collected a sample of their credit card accounts over the past year. The accounts include the current balance on the account, the annual income of the account holder and whether the account holder was a student (coded 1: if a student and 0 if not). They also indicated whether the account was in default (coded 1: if default and 0 if not). They are interested in knowing whether bank balance and/or student status contributes to the probability of a default. The file Credit_Card_Defaults.sav provided information about subsample. Variable definitions are provided in the values and Labels attributes as well. (100 points)

1. Credit Card Defaults: 1. Run the appropriate analysis to answer the following (based on the sample): i. What is the sample size?
2. ii. How many/What Percentage of credit card holders are students/non-students?
3. iii. How many/What Percentage of customers are in default/in good standing?

1. 2. Run the appropriate descriptive for Income and Account Balance. You should be able to answer average/minimum/maximum/95% confidence interval/ Standard deviation of Income and/or Account balance of 1. Customers (includes students, non-students, defaulters and the ones in good standing)
2. 2. students and non-students
3. 3. defaulters and the non-defaulters

1. 3. You know that generally banks maintain their customer base such that about 30% are students (students are considered riskier, so the bank wants to manage the investment risk by not granting too many student credit cards). Is your bank at par with the industry practices (in terms of student status)? Run the appropriate statistical test and be prepared to provide statistical justification. You may be asked to provide the p-value of the test.
2. 4. The industry benchmark for credit card default is 3% (i.e. on an average, 3% of all credit card holders are in default status). Is your bank at par with the industry (in terms of % of defaulters?)? Run the appropriate statistical test and be prepared to provide statistical justification. You may be asked to provide the p-value of the test.
3. 5. What percentage of students are in default/good standing? What percentage of non-students are in default/good standing?

1. i. What is the probability that a student is in default? If a customer is in default what is the probability that he/she is a student?
2. ii. Run the appropriate statistical test to check whether the proportion of students who default is any different that the proportion of non-students who are in default. You may be asked to provide the p-value of the test.
3. iii. What does your analysis say about the odds of default by students vs. non-students? If your analysis shows that students are more likely to default, think about some business steps you can take. On the other hand, if your analysis shows that students are not different than non-students in terms of likelihood to default, think about some possible business strategies as well.