Most studies are based on data that are recorded with some errors. We may be able to record gender precisely, birth weight with some errors, birth length with more errors, past occupational exposures with even more uncertainties, not to mention long-term dietary habits or use of mobile phones years ago. Sometimes diseases are diagnosed by their symptoms only, like depression. Sometimes there may be objective signs, as for hypertension. In clinical practice, there will often be some disagreement between groups of clinical experts when making diagnoses.