Claim That Correcting 'Two Errors' in a South Korean Study Restores Clear Evidence for Cosmic Acceleration: Unverifiable and Likely Backwards

The claim holds that when two specific errors in a South Korean cosmology study are corrected, the supernova evidence for cosmic acceleration comes back strong and consistent. The verdict is unverifiable as stated — and the framing almost certainly misrepresents the source it is drawing on.

The study most plausibly being referenced is Colin, Mohayaee, Rameez, and Sarkar (2019), published in Astronomy & Astrophysics 631, L13, which includes South Korean-affiliated authors and analyzed 740 Type Ia supernovae from the JLA dataset. That paper found a roughly 2.8σ dipole anisotropy in the deceleration parameter, suggesting the acceleration signal may partly reflect a local bulk-flow artifact rather than a genuine cosmological constant. A follow-up by Rameez and Sarkar (2021), published in Classical and Quantum Gravity 38, 154005, argued explicitly that after correcting for two biases — selection effects and local bulk flow — the isotropic acceleration signal drops to approximately 2σ. Mohayaee, Rameez, and Sarkar (2021) in the European Physical Journal Special Topics state outright that after these corrections, the evidence for acceleration 'is not statistically significant' from supernovae alone. The authors' own conclusion is the opposite of what the claim describes.

To steelman the claim: one could argue that the Sarkar group's corrections expose flaws in earlier analyses, and that a properly rebuilt analysis on clean data would vindicate acceleration. That is a coherent position — but no published peer-reviewed study actually demonstrates it using the specific 'two errors' framing. The claim presents an outcome that does not appear in the literature it presumably cites. That makes it unverifiable, not merely contested.

The mainstream rebuttal to the Sarkar group's methodology is direct and well-sourced. Rubin and Hayden (2016), in Astrophysical Journal Letters 833, L30, responded to related work by Nielsen et al. (2016) — itself sometimes conflated with the South Korean study — showing that correcting for selection effects and using a proper likelihood actually recovers greater than 5σ evidence for acceleration from the same Union2.1 dataset. Tutusaus et al. (2017), in Astronomy & Astrophysics 602, A73, confirmed that combining supernovae with CMB and BAO data keeps the acceleration signal above 5σ, while also honestly noting that supernovae alone yield weaker standalone constraints. Most decisively, Brout et al. (2022) — the Pantheon+ analysis in Astrophysical Journal 938, 110 — used 1,701 supernova light curves and found Ω_Λ = 0.334 ± 0.018 with greater than 5σ evidence for acceleration when combined with CMB data. The Sarkar group's specific corrections are not adopted anywhere in that analysis.

What is genuinely true here deserves acknowledgment: the Sarkar group raises real methodological questions about selection bias and bulk-flow contamination in supernova surveys, and the debate over whether supernovae alone constitute iron-clad proof of acceleration is legitimate minority science, not fringe crankery. Nielsen et al. (2016) in Scientific Reports 6, 35596 found only roughly 3σ evidence using a model-independent approach on the same Union2.1 data — a result that was published in a peer-reviewed journal and is worth taking seriously. The disagreement is about methodology, not fabrication.

The manipulation pattern here is a rhetorical inversion: take a real scientific dispute, swap the direction of the conclusion, and present it as a clean empirical fix. 'Correcting errors restores the truth' sounds like good-faith science, but when the researchers who identified the errors say those corrections eliminate the signal — not restore it — the framing has been reversed. Watch for claims that cite a contested minority study while describing its conclusion as the opposite of what the authors actually wrote.