Patricio CuarĂ³n schrieb:
 Is it correct that ADF always uses the whole range instead of the
 current sample?
  
I don't see that behavior here.
 
 
     ? smpl
 
     Full data range: 2002:1 - 2008:4 (n = 28)
 
 
     ? smpl
 
     Full data range: 2002:1 - 2008:4 (n = 28)
 
     Current sample: 2004:1 - 2008:4 (n = 20)
  
here you're not showing what happened in between right?
 
     ? adf 8 log(salarios) --ct
 
  
My guess is that gretl's adf routine allows the initial values to be
taken from the pre-sample periods, if available. Since you have 8 lags
and by coincidence shorten the sample also by 8 obs, it looks as if
gretl is always using the full sample, but it really doesn't.
cheers,
sven