Accuracy

[ams_nanolab]

Say I have to do a calculation with a very large number of atoms and I am not doing scf calculations.
Which will be more accurate: non-scf Slater-Koster or non-scf Extended Huckel method. ?

[Anders Blom]

Impossible to answer such a question because the term "accuracy" is not unambiguous. If you are considering SK and EHT, the main question is about the quality of the parameters, so it can only be answered by comparing to DFT and/or experiments. But you can of course perform that comparison for a small system.