dwaine <- read.table("http://www.stat.ufl.edu/~winner/sta4210/data/dwaine_studios.txt", header=F,col.names=c("Y","X1","X2")) attach(dwaine) dwaine n <- length(Y) Y_corr <- (1/sqrt(n-1))*(Y-mean(Y))/sd(Y) # Correlation transformation for Y X1_corr <- (1/sqrt(n-1))*(X1-mean(X1))/sd(X1) # Correlation transformation for X1 X2_corr <- (1/sqrt(n-1))*(X2-mean(X2))/sd(X2) # Correlation transformation for X2 print(cbind(Y_corr,X1_corr,X2_corr)) dwaine.stdreg <- lm(Y_corr ~ X1_corr + X2_corr -1) # Regression of Y* on X1*,X2*, no intercept summary(dwaine.stdreg) (b1 <- (sd(Y)/sd(X1))*coef(dwaine.stdreg)[1]) # Compute b1 from b1* (b2 <- (sd(Y)/sd(X2))*coef(dwaine.stdreg)[2]) # Compute b2 from b2* (b0 <- mean(Y) - b1*mean(X1) - b2*mean(X2)) # Comute b0 dwaine.reg <- lm(Y ~ X1 + X2) # Regression of Y on X1,X2 summary(dwaine.reg)