Power-Spectrum Analysis of Reconstructed DAMA Data

TL;DR

This paper independently analyzes DAMA's reconstructed data, confirming an annual oscillation with a peak in early June, but with reduced statistical significance, and explores different analysis methods and error treatments.

Contribution

It provides an independent reanalysis of DAMA data using multiple statistical methods, highlighting the impact of error treatment and data binning on significance.

Findings

Confirmed annual oscillation with early June peak

Lower significance level than original DAMA claims

Explored effects of error treatment and binning

Abstract

Claims by the DAMA (DArk MAtter) collaboration to have detected an annually varying signal consistent with models of dark matter appear to be at variance with results from other dark-matter searches. To further understand the DAMA results, we have carried out an independent analysis of DAMA data reconstructed from published figures. In addition to reexamining the Lomb-Scargle and chi-square analyses previously carried out by the DAMA collaboration, we carry out two new likelihood analyses and a new chi-square analysis, focusing attention on the treatment of experimental errors and binning. We confirm the existence of an annual oscillation, with a maximum in early June, but at a lower significance level than previously reported.