A quantitative comparison between an experimental and a theoretical description of any Auger spectrum requires a common energy reference point such as the Fermi level. With non-conductors this is problematic, since the Fermi level is not well defined and charging may shift the Fermi level. A common approach to overcome this problem employs spectra containing the Auger peak, core, and valence band (VB) photoemission peaks. One then has only to locate a common reference point in the Auger and VB spectra relative to the core level peak. We have developed a procedure where a comparison is made between model theoretical and experimental spectra so that a large energy range (about 20 eV) of each spectrum can be utilized to locate the common reference points. We have found that this method enables accurate reference points to be determined even with severe sample charging. The method is precise enough to observe what may be a differential charging effect in the polyethylene (PE) spectra. This suggests that the photoelectron and Auger peaks may not shift an equal amount with charging in all instance. Although much smaller than that previously reported, a breakdown in the Cini-Sawatzky expression as modified for covalent systems still is suggested for PE.