Abstract
It is shown how error estimates for intensities derived from area detector data in powder diffraction experiments should be calculated. It is demonstrated that the uncritical application of standard weighting schemes to the derived one dimensional data in subsequent least-squares refinements results in meaningless goodness of fit values less than unity that prohibit the evaluation of the quality of the fit. Moreover, it is shown that the calculated estimated standard deviations of the fitted parameters can only be calculated correctly when the true uncertainties of the measured intensities are known.