Theoretical vs. empirical discriminability: the application of ROC methods to eyewitness identification

摘要: Receiver operating characteristic (ROC) analysis was introduced to the field of eyewitness identification 5 years ago. Since that time, it has been both influential and controversial, debate raised an issue about measuring discriminability is rarely considered. The concerns distinction between empirical (measured by area under ROC curve) vs. underlying/theoretical d’ or variants it). Under most circumstances, two measures will agree a difference conditions in terms discriminability. However, possible for them disagree, fact can lead confusion which condition actually yields higher For example, if have implications real-world practice (e.g., comparison competing lineup formats), should policymaker rely on area-under-the-curve measure theory-based measure? Here, we illustrate given as many underlying there are theories one willing take seriously. No matter theory correct, practical purposes, singular best identifies diagnostically superior procedure. reason, curve informs policy way theoretical never can. At same equally important, but different reason. Without adequate understanding relevant task, be no position enhance