The number of good reflections in a powder pattern

The number of `statistically independent' observations in a powder diffraction pattern has been the topic of some discussion over the past few years. It is argued that while this notion is tantalizing, it is illusory in principle, and it is suggested that a more appropriate measure of the quality of the data is the number of `good' reflections. By way of explanation, a tutorial on the underlying concepts of correlation and covariance, which are a surprisingly common source of confusion, is provided. A practical procedure for implementing the theoretical ideas is proposed.