[Ifeffit] Artemis: Rejects first shell fitting

Hello all, I am trying to fit my data using Artemis for a reference foil sample. However, the fit seems to get worse when I include the scattering paths at lowest R (i.e. the first shell scattering, with the highest rank). Does it make physical sense to not include this in the fitting? Is it acceptable to try and fit only a certain range of the |X(R)| vs R for the foil reference - in order to estimate the amplitude reduction factor for subsequent fitting of the real samples? Any insight/advice is much appreciated. Thank you. Best, A