*Set directory to where the BlankWithId500.dta file is located:
cd "C:\Documents and Settings\Administrator\My Documents\Teaching\581\Fall2008\classData\Simulated\"
use "BlankWithId500.dta", clear

gen x = rnormal(3,1)
gen D = rbinomial(1, .6)

*Select one of the following generation statements by moving the star from one
*statement to the other. Feel free to alter the coefficients.
gen y = 5 + 4*x + 2*D + 3*x*D + rnormal(0,3)
*gen y = 5 + 4*x + 2*D + rnormal(0,3)

gen xD=x*D
label var xD "interaction term"
*Notice that the correlation between the x and D variables are fine while 
*the correlation between x and xD or between D and xD are high.
corr x D xD

regress y x D 
regress y x D xD
predict yhat0 if D==0
label var yhat0 "D=0"
predict yhat1 if D==1
label var yhat1 "D=1"
scatter y x || connected yhat0 x, lcolor(red) msymbol(i) || connected yhat1 x, lcolor(blue) msymbol(i)