Provide the Relevant Outputs Produced by R - Management Assignment Help

SECTION B. Provide the relevant outputs produced by R (you can copy/paste or screenshot your relevant outputs). There is no need to provide the R codes. Your solutions must be very detailed with clear interpretations. Round your answers to the nearest 4 decimals.

1.  Conduct the appropriate t-test (ignore the fact that we have a large sample size) to see whether married female have higher credit limit than not-married male. Assume variances for both groups are equal. Use ???? = 0.05. [HINT: pay attention to male when sub-setting (filtering)]
   1. What are the averages and standard deviations of both samples’ credit limits?
   2. State your hypotheses. Indicate which test you are using.
   3. Provide your test statistic, rejection region and critical value.
   4. Provide your decision and interpretations.

2. Create a multiple linear regression model to predict credit limit of married female individuals [HINT: subset this dataset before proceeding with your solutions] based on the following explanatory variables: Income, Rating, Balance and Ethnicity. Make sure to have Caucasian as the base (reference) group. Use a = 0.01 for significance tests.
   1. Provide your fitted model after running the regression
   2. Interpret the partial regression coefficients of the model?
   3. Provide your fitted models for Asian, African American, and Caucasian individuals based on your model in part a.  

3. Conduct an appropriate test to see whether Ethnicity has any significant impact on Income. Use a = 0.06.
   1. Provide your hypotheses. Which statistical test is the most appropriate?
   2. List all the assumptions and conditions of the test.
   3. Provide your test statistic, rejection region and critical values.
   4. Provide your decision and conclusion.

4. Conduct an appropriate test to see whether being a student and Ethnicity among individuals with at least 3 cards have any significant association. Use a = 0.10.
   1. Provide your hypotheses. Which statistical test is the most appropriate?
   2. Provide your actual, expected tables and test statistic value.
   3. Provide the rejection region and critical values.
   4. Check the validity of this test.
   5. Provide your decision and conclusion.

5.  Fit a Multiple Logit Model to predict whether a customer paid back all credit balance (1) or did not pay back (0) using the following independent variables: Age, Gender, Income, Ethnicity and Education.
   1. Write down your fitted logit model.
   2. Compute the odds ratio and probability that the customer paid back the credit balance for a 25-year-old Asian male, who has 12 years of education and an income of $50000.
   3. Provide your Confusion Matrix by comparing your actual values with predicted values. Calculate Accuracy, Sensitivity, Specificity and Precision ratios and Interpret all ratios. Use all your observations (do not split your data into train and test sets) to compute these ratios.