hw8 - Yue Wang HW8 Graduate Problem 1 1 >...

This preview shows page 1 - 3 out of 8 pages.

Yue Wang , HW8, GraduateProblem 1:1)> g1<-glm(colonies~log(dose+1),family=poisson)> summary(g1)Coefficients:Estimate Std. Error z value Pr(>|z|) (Intercept) 3.01989 0.09712 31.095 < 2e-16 ***log(dose + 1) 0.08585 0.02018 4.255 2.09e-05 ***Null deviance: 78.358 on 17 degrees of freedomResidual deviance: 59.629 on 16 degrees of freedomAIC: 156.17Number of Fisher Scoring iterations: 4If we don’t take log(dose+1), the coefficient is not significant.> g1<-glm(colonies~dose,family=poisson)> summary(g1)Coefficients:Estimate Std. Error z value Pr(>|z|) (Intercept) 3.3219950 0.0540292 61.485 <2e-16 ***dose 0.0001901 0.0001172 1.622 0.105 Null deviance: 78.358 on 17 degrees of freedomResidual deviance: 75.806 on 16 degrees of freedomAIC: 172.34Number of Fisher Scoring iterations: 42)> pchisq(deviance(g1),df.residual(g1),lower=FALSE)[1] 6.042092e-07p-value is too small, indicating a lack-of-fit problem.3) From the residual vs. fitted plot, the line is pretty flat except that there are two outliers: #12 and #16. Normal Q-Q plot is pretty a straight line. Scale location plot is
Yue Wang , HW8, Graduate

  • Left Quote Icon

    Student Picture

  • Left Quote Icon

    Student Picture

  • Left Quote Icon

    Student Picture