Fifty years of primordial helium abundances: A statistical reanalysis

Abstract

The primordial helium mass fraction, $Y_\mathrm{p}$, is a key observational pillar of Big Bang nucleosynthesis and a sensitive probe of early-Universe physics. Over the past several decades, numerous observational $Y_\mathrm{p}$ determinations have been published using a wide range of astrophysical tracers and cosmological techniques. Although recent measurements exhibit striking convergence and increasingly small uncertainties, the statistical and historical context of this consensus has not been examined systematically. Here, we compile and analyse a comprehensive dataset of observational $Y_\mathrm{p}$ determinations published between the late-1960s and 2022. The final sample comprises 143 reported values spanning multiple tracers. We find clear evidence for long-term convergence in published $Y_\mathrm{p}$ values, punctuated by statistically significant change points in the mid-2000s and early 2010s. Careful examination reveals that many extragalactic H{\sc ii}-region determinations are not fully independent, relying on re-analyses or partial reuse of a limited number of observational datasets. This reduces the effective number of independent constraints and provides important context for interpreting the precision of recent results. Our findings do not challenge the overall consistency of modern $Y_\mathrm{p}$ determinations with standard cosmology, but they underscore the importance of accounting for data dependence, methodological homogeneity and historical evolution when synthesising measurements.