> logit_1 summary(logit_1) Call: glm(formula = DEFAULT ~ RATING + SIZE + LLPLOANS, family = binomial, data = ratings) Deviance Residuals: Min 1Q Median 3Q

> logit_1 <- glm(formula = DEFAULT ~ RATING + SIZE + LLPLOANS, family = "binomial", data = ratings)

> summary(logit_1)

Call:

glm(formula = DEFAULT ~ RATING + SIZE + LLPLOANS, family = "binomial",

data = ratings)

Deviance Residuals:

Min 1Q Median 3Q Max

-1.5486 -0.8492 -0.6274 1.1581 2.6677

Coefficients:

Estimate Std. Error z value Pr(>|z|)

1. -7.8017 2.6862 -2.904 0.00368 **

RATING"2" 2.0454 1.0491 1.950 0.05122 .

RATING"3" 3.6292 1.1075 3.277 0.00105 **

SIZE 0.2596 0.1310 1.982 0.04749 *

LLPLOANS -30.8333 19.7686 -1.560 0.11883

---

Signif. codes: 0 *** 0.001 ** 0.01 * 0.05 . 0.1 1

(Dispersion parameter for binomial family taken to be 1)

Null deviance: 238.79 on 196 degrees of freedom

Residual deviance: 213.64 on 192 degrees of freedom

AIC: 223.64

Number of Fisher Scoring iterations: 5

2. > logit_2 <- glm(formula = DEFAULT ~ RATING + SIZE + LLPLOANS + ROE, family = "binomial", data = ratings)

> summary (logit_2)

Call:

glm(formula = DEFAULT ~ RATING + SIZE + LLPLOANS + ROE, family = "binomial",

data = ratings)

Deviance Residuals:

Min 1Q Median 3Q Max

-2.0232 -0.7590 -0.5145 0.8081 2.3591

Coefficients:

Estimate Std. Error z value Pr(>|z|)

1. -5.7195 2.7426 -2.085 0.03703 *

RATING"2" 1.5603 1.0641 1.466 0.14255

RATING"3" 3.2671 1.1174 2.924 0.00346 **

SIZE 0.2048 0.1347 1.521 0.12829

LLPLOANS -41.1917 27.2973 -1.509 0.13130

ROE -8.2431 2.0702 -3.982 6.84e-05 ***

---

Signif. codes: 0 *** 0.001 ** 0.01 * 0.05 . 0.1 1

(Dispersion parameter for binomial family taken to be 1)

Null deviance: 238.79 on 196 degrees of freedom

Residual deviance: 191.79 on 191 degrees of freedom

AIC: 203.79

Number of Fisher Scoring iterations: 5

1. Interpret results of logit_1 and logit_2

Compare the results, and explain the difference. What does the difference mean?