A bank manager ran a logistic regression to examine whether and how a borrower's age (in years), household income (in thousands), debt to income ratio, and credit card debt (in thousands) affect whether the borrower will commit debt default (=1) or not (=0). The following are the logistic regression results (α = 0.05).

Classification Tablea
	Observed		Predicted
			Previously defaulted		Percentage Correct
			No	Yes
Step 1	Previously defaulted	No	3554	190	94.9
		Yes	856	400	31.8
	Overall Percentage				79.1
a. The cut value is .500

 

Variables in the Equation
		B	S.E.	Wald	df	Sig.	Exp(B)
Step 1a	Age in years	-.069	.006	142.064	1	.000	.933
	Household income in thousands	-.014	.002	43.083	1	.000	.986
	Debt to income ratio (x100)	.093	.007	188.459	1	.000	1.097
	Credit card debt in thousands	.328	.031	112.888	1	.000	1.389
	Constant	.276	.186	2.189	1	.139	1.317
a. Variable(s) entered on step 1: Age in years, Household income in thousands, Debt to income ratio (x100), Credit card debt in thousands.

1.) Assuming that a potential borrower is 30 years old with $80,000 household income, debt to income ratio of 12, credit card debt of $2,000, which is his/her probability of committing debt default?

 

2.) For every $1000 increase in income, the odds of a borrower committing debt default vs. not committing debt default ___________

 

3.) With the default 0.5 cutoff, what percentage of observations were correctly predicted by the manager's model?