False Rejection Rate (FRR, Type I error)

The false rejection rate measures how often a biometric system fails to match a legitimate user’s sample to their own enrolled template and refuses them. Study sources commonly call a false rejection a Type I error. NIST SP 800-63B-4 uses the term false non-match rate and observes that false non-matches can often be mitigated by repeating the measurement or using another authentication method.

The two error rates pull against each other. Making the system stricter lowers the false acceptance rate and raises FRR, and the point where they are equal is the crossover error rate. A false rejection admits no one who should be kept out, but a high FRR has real costs: users are delayed and lose trust in the system. An organisation frustrated by rejections may relax the threshold or add a weak fallback, which brings the risk back by another route. User acceptance is therefore part of choosing a biometric system.

Exam relevance: a scenario is likely to describe an authorised employee repeatedly refused by a reader and ask which error it shows, where false rejection, Type I, is the expected label. Candidates are expected to keep the Type I and Type II labels the right way round, and to see that a high-security area commonly accepts a higher FRR as the price of a lower FAR.