Data correlation analysis applied to electron diffraction

Publisher Website
Google Scholar
Abstract
A method of data evaluation is proposed for electron diffraction data whereby nonreproducible systematic errors are taken into account in the determination of the uncertainties. Artificial, computer generated, statistical noise is added to the data. The extent to which noise must be added is evaluated by the correlation coefficient function. Such a degree of noise allows the total standard deviation of the derived structure parameters to be treated as a Gaussian distribution. This method is illustrated by eight sets of data collected for SF6 at electron energies of 54 keV on a counting electron diffraction unit. It is shown that, by addition of a maximum of 0.3% noise to the data, correlation was reduced to insignificance.