Very good lecture videos which I used when I was a beginner (for the links to the vidoes you have to scroll down): http://www.diamond.ac.uk/Beamlines/Spectroscopy/Techniques/XAS.html
Finally, I would suggest to talk to people that understand the technique every time you get a fit to the spectrum/spectra. For example, what I still tend to do is check with the beamline scientists at the Synchrotron where the data was collected if the fit makes sense, and if I'm not using too many parameters (or too few) in the fit.
One other tip I can give you is to first think what you think might be the crystal structure and check this on the ICSD website what kind of interatomic distances you'd expect, then use the crystallographic information file to calculate the single scattering paths AND the multiple scattering paths (they can be very important too) and only then try and see if your data matches your prediction. And use this iteratively to get to the final answer. I know this is not always possible, but at least this way you can already exclude a lot of possibilities.