For data introduced in Exercise 4.10 about AIDS symptoms, AZT use, and race, here is some R output:

-------------------------------------–-------------------------------------

> fit <- glm(yes/(yes+no) ~ azt + race, weights=yes+no, family=binomial,

+ data=AIDS)

> summary(fit)

Deviance Residuals:

1234

-0.5547 0.4253 0.7035 -0.6326 Estimate Std. Error z value Pr(>|z|)
(Intercept) -1.07357 0.26294 -4.083 4.45e-05 aztyes -0.71946 0.27898 -2.579 0.00991 racewhite 0.05548 0.28861 0.192 0.84755

---

Null deviance: 8.3499 on 3 degrees of freedom Residual deviance: 1.3835 on 1 degrees of freedom

> 1 - pchisq(1.3835, 1)

[1] 0.23950

> cbind(AIDS$azt, AIDS$race, fitted(fit), rstandard(fit,type="pearson"), residuals(fit,type="pearson"), residuals(fit,type="deviance"))

[,1] [,2] [,3] [,4] [,5] [,6]

1 2 2 0.1496 -1.1794 -0.5447 -0.5547 # azt=yes, race=white 2 1 2 0.2654 1.1794 0.4282 0.4253 # azt=no, race=white 3 2 1 0.1427 1.1794 0.7239 0.7035 # azt=yes, race=black 4 1 1 0.2547 -1.1794 -0.6220 -0.6326 # azt=no, race=black

-------------------------------------–--------------------------------------

a. Test the model goodness of fit and interpret the result.

b. Explain how the relative sizes of the fitted values reflect the results of the individual tests for the AZT effect and the race effect.

c. The display shows the standardized residuals, Pearson residuals, and deviance residuals. Explain advantages of using standardized residuals rather than the others.