SLID1 <- read.table("c:/temp/SLID-Ontario.txt",sep = "\t",header=F,skip=1)
names(SLID1)  <- c("age","sex","wages","yearsed")

SLID1$male <- as.numeric(SLID1$sex == "Male")

 # ----------- fit the model on the original scale -----------------------------

SLID1$log2wages <- log2(SLID1$wages)

SLID1.lm1 <- lm(wages ~ male +age +yearsed, data=SLID1)

 # automatically generates several useful plots
par(mfrow=c(2,2)) 
plot(SLID1.lm1)
par(mfrow=c(1,1)) 

 # A kernel density estimate of the residuals
SLID1dens <- density(rstudent(SLID1.lm1))

 # A plot of the histogram and kernel density estimate
hist(rstudent(SLID1.lm1),prob=T,ylim=c(0,.46),main="Distribution of studentized residuals")
lines(SLID1dens)


 # ---------- now try the log base 2 data ---------------------------------------

SLID1.lm2 <- lm(log2wages ~ male +age +yearsed, data=SLID1)

 # automatically generates several useful plots
par(mfrow=c(2,2)) 
plot(SLID1.lm2)
par(mfrow=c(1,1)) 

 # A kernel density estimate of the residuals
SLID1dens2 <- density(rstudent(SLID1.lm2))

 # A plot of the histogram and kernel density estimate
hist(rstudent(SLID1.lm2),prob=T,ylim=c(0,.46),
           main="Distribution of studentized residuals for log data")
lines(SLID1dens2)


 # remove objects when finished

rm(SLID1.lm1,SLID1dens,SLID1.lm2,SLID1dens2)