Statistical methods must be described in sufficient detail to enable a knowledgeable reader with access to the original data to verify the reported results. Whenever possible present them with appropriate indicators of measurement error or uncertainty.