The Profiled Feldman-Cousins method for confidence interval estimation in the presence of nuisance parameters

Determining statistically accurate confidence intervals for experiments with small datasets and bounded parameters is a challenge. The Unified Approach proposes a solution to address these issues but does not offer a prescription on how to handle nuisance parameters. Several approaches were considered, and the Profiled Feldman-Cousins method presented in this poster was found to provide the most accurate statistical coverage. Its implementation in the context of NOvA's neutrino oscillation measurement is also illustrated here. We believe this method could be effective in other experiments sharing similar challenges.