error costs

Different kinds of error often incur different costs. For example when diagnosing a disease a false positive might lead to additional costs of tests or unnecessary treatment and distress for the patient, whereas a false negative might mean that a potentially fatal or debilitating condition is missed.

In decision making empirical evidence is often combined with knowledge (or estimates) of the base rate to estimate the probability of different forms of error and balance this with the differing costs. This may be calculated exactly using Bayesian methods or assessed more informally.