Reliability Versus Validity of Tests Used in Genetic Identification

摘要

An ever-increasing use of genetic traits for the identification of suspects and evidence in criminal cases has brought a corresponding increase in challenges to the use of the test results. The ensuing arguments have, unfortunately, produced a blurring of two essential aspects of any test, namely, its reliability versus its validity. Reliability of a test refers to the consistency or reproducibility of the test results under a set of prescribed conditions. Validity refers to whether the test actually measures or specifies the trait in question. Experts in a field may generally agree that a test has a very high reliability but may seriously disagree as to its validity. IQ tests are the prime examples of this disagreement among experts. Whereas there is little debate that the tests give consistent and reproducible results, there is a serious question raised as to whether such tests actually measure the intelligence of individuals.