Diagnostic plots for berkeley data

Prepare data frame for plotting

berkeley <- as.data.frame(UCBAdmissions)
cellID <- paste(berkeley$Dept, substr(berkeley$Gender,1,1), '-', 
                substr(berkeley$Admit,1,3), sep="")
rownames(berkeley) <- cellID

using glm()

berk.mod <- glm(Freq ~ Dept * (Gender+Admit), data=berkeley, family="poisson")
summary(berk.mod)

## 
## Call:
## glm(formula = Freq ~ Dept * (Gender + Admit), family = "poisson", 
##     data = berkeley)
## 
## Deviance Residuals: 
##     Min       1Q   Median       3Q      Max  
## -3.4776  -0.4144   0.0098   0.3089   2.2321  
## 
## Coefficients:
##                     Estimate Std. Error z value Pr(>|z|)    
## (Intercept)          6.27557    0.04248 147.744  < 2e-16 ***
## DeptB               -0.40575    0.06770  -5.993 2.06e-09 ***
## DeptC               -1.53939    0.08305 -18.536  < 2e-16 ***
## DeptD               -1.32234    0.08159 -16.207  < 2e-16 ***
## DeptE               -2.40277    0.11014 -21.816  < 2e-16 ***
## DeptF               -3.09624    0.15756 -19.652  < 2e-16 ***
## GenderFemale        -2.03325    0.10233 -19.870  < 2e-16 ***
## AdmitRejected       -0.59346    0.06838  -8.679  < 2e-16 ***
## DeptB:GenderFemale  -1.07581    0.22860  -4.706 2.52e-06 ***
## DeptC:GenderFemale   2.63462    0.12343  21.345  < 2e-16 ***
## DeptD:GenderFemale   1.92709    0.12464  15.461  < 2e-16 ***
## DeptE:GenderFemale   2.75479    0.13510  20.391  < 2e-16 ***
## DeptF:GenderFemale   1.94356    0.12683  15.325  < 2e-16 ***
## DeptB:AdmitRejected  0.05059    0.10968   0.461    0.645    
## DeptC:AdmitRejected  1.20915    0.09726  12.432  < 2e-16 ***
## DeptD:AdmitRejected  1.25833    0.10152  12.395  < 2e-16 ***
## DeptE:AdmitRejected  1.68296    0.11733  14.343  < 2e-16 ***
## DeptF:AdmitRejected  3.26911    0.16707  19.567  < 2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## (Dispersion parameter for poisson family taken to be 1)
## 
##     Null deviance: 2650.095  on 23  degrees of freedom
## Residual deviance:   21.736  on  6  degrees of freedom
## AIC: 216.8
## 
## Number of Fisher Scoring iterations: 4

Influence plot

influencePlot(berk.mod, id=list(n=3, labels=cellID))

##           StudRes       Hat      CookD
## AM-Adm -4.1541239 0.9588091 22.3046668
## AM-Rej  4.1497537 0.9254346 11.8924974
## AF-Adm  4.0991865 0.6853494  2.0871343
## AF-Rej -4.4178464 0.4304068  0.7240710
## BM-Adm -0.5037122 0.9842940  0.8833704
## BM-Rej  0.5036947 0.9729710  0.5074049
## FM-Rej  0.6197342 0.9692308  0.6721646

op <- par(mfrow = c(2,2))
plot(berk.mod)

par(op)