Statistical comparison of two methods of clinical measurement

The comparison of two measurement techniques, neither of which can be regarded as being a 'gold standard', is fundamental to much basic research. This apparently simple problem raises many statistical difficulties, which are reviewed in this paper in the context of anorectal manometry. It is concluded that a recently developed approach focusing on the agreement between the two techniques is far superior to the more conventional approach based on correlation. However, until the benefits of the new analysis are more widely recognized (and it is hoped that this paper is a step towards that goal), it is argued that it is useful to present both analyses in such comparisons of measurement techniques.