basic question about ECM and Cointegrating regression

[thwiaethae]

Hello, I have some questions about ECM and cointegrating regression. (maybe very basic...)

1. I want to know the single equation error correction method which may be estimated using the Equation object and conventional OLS routines.

<step1> Find variables related to cointegration using cointegration test. (eg. y1, y2).
<step2> regress y1(t) on c y2(t) by OLS. AND GET error correction term e(t)=y1(t)-c-beta*y2(t)
<step3> regress d(y1(t)) on e(t-1) c d(y2(t)).

Is that right? If is there any error what I know, please let me know.

2. In DOLS, FMOLS, CCR
1) I(1) variable related cointegration
2) I(1) variable not related cointegration
3) I(0) variable

I want to know how 1)~3) should be handled in equation (differential??, Equation specification??)