Sequential Analysis Techniques for Correlation Studies in Particle Astronomy

TL;DR

This paper introduces a sequential analysis method for detecting and evaluating correlations between cosmic ray arrival directions and astrophysical objects, allowing continuous monitoring and rigorous significance assessment.

Contribution

It presents a novel sequential testing approach that accounts for unknown signal strength and enables ongoing analysis in particle astronomy studies.

Findings

Method allows real-time correlation detection.

Accounts for unknown signal strength.

Supports post-hoc significance evaluation.

Abstract

Searches for statistically significant correlations between arrival directions of ultra-high energy cosmic rays and classes of astrophysical objects are common in astroparticle physics. We present a method to test potential correlation signals of a priori unknown strength and evaluate their statistical significance sequentially, i.e., after each incoming new event in a running experiment. The method can be applied to data taken after the test has concluded, allowing for further monitoring of the signal significance. It adheres to the likelihood principle and rigorously accounts for our ignorance of the signal strength.