A Meta-Analysis Of LHC Results

TL;DR

This paper conducts a meta-analysis of LHC experiment results to identify potential biases in published uncertainty values, using two independent graphical analysis techniques on exclusion plots.

Contribution

It introduces a novel dual-method approach to assess the accuracy of uncertainty estimates in high-energy physics publications.

Findings

Analysis of exclusion plots shows deviations from expected statistical distributions.

The study identifies potential biases in published uncertainty values.

Results suggest the need for improved reporting standards in physics experiments.

Abstract

We report the preliminary results of a meta-analysis conducted to examine possible biases in the uncertainty values published in papers by ATLAS and CMS experiments. We have performed this analysis using two independent techniques; a vectoral analysis of the vector graphics files and a bitmap analysis of the raster graphic files of the exclusion plots from various physics searches. In both procedures, the aim is to compute the percentages of the data points scattered within 1-sigma and 2-sigma bands of the plots and verify whether the measured percentages agree with statistical norms assuming unbiased estimations of the uncertainties.