By estimation of p by Durbin-Watson d statistics,
Then convert both x and y into x* and y* respectively.
And type the followings on the command line to generate x* and y*
"genr xstar = x - ( 0.312981 * x(-1))"
"genr ystar = y - ( 0.312981 * y(-1))"
And then regress ystar on xstar to examine the results
And the results are as follows:
