General to specific technique

[27npg]

Hi - I am building a structural model of a South American economy. The model will consist of 12 or so OLS estimations which will then run as SEM in the model function of Eviews. I am trying to fit the best individual equations and the literature has discussed a general to specific modelling technique to get the best fit re number of lags and relevant variables in each equation. Does anyone know the best technique to use to do this in Eviews.

Thanks in advance

NPG