Chemistry with confidence: should Clinical Chemistry require confidence intervals for analytical and other data?

Abstract Confidence intervals are not commonly provided with analytical or other data reported in Clinical Chemistry although P values are. However, confidence intervals provide an explicit demonstration of the direction and magnitude of uncertainty and are intuitively easy to grasp, unlike P values. It is therefore argued that the Journal should adopt a policy requiring the provision of confidence intervals. Such a policy would improve the statistical rigor of Journal reports.