Abstract
The reliability of the partial structure factors determined by either of two x-ray scattering techniques using the effects of anomalous dispersion is considered. A comparison of the two experimental techniques is given for both binary and ternary systems. The recently proposed x-ray frequency-modulation technique is found to be about an order of magnitude better than the direct x-ray anomalous scattering method, and it is also found to be suitable for ternary systems. Experimental error is simulated by a pseudo-random-number generator which produces normally distributed numbers with a specified mean and standard deviation. Conditions corresponding to about 1% experimental error from data acquisition and processing are assumed.