Blind unblinding procedure for the PADME X17 data sample

Abstract

The PADME experiment at the Frascati DA$\Phi$NE LINAC has performed a search for the hypothetical X17 particle, with a mass of around 17 MeV, by scanning the energy of a positron beam striking a fixed target. The X17 should be produced from the resulting $e^+e^-$ annihilation. Since the expected mass of this particle is only roughly known, data sidebands cannot be clearly defined. Furthermore, the need to keep the analysis blind to potentially sizable signal contributions prevents a clear assessment even of the quality of the data sample in this search. In light of these challenges, this paper presents an innovative strategy adopted by the PADME Collaboration to perform data quality checks without disclosing the X17 sample. Moreover, the procedure designed to eventually unblind the data is described, together with the statistical approach adopted to extract the limits on the coupling between the X17 and the Standard Model.