Overview
- Led by Sergey Frolov with colleagues in Minnesota and Grenoble, the team replicated four high-profile signatures in nanoscale devices using non-topological mechanisms.
- The work consolidates multiple replication attempts into a single paper that Science published after a two-year peer and editorial review.
- The authors say earlier follow-ups were rejected by leading journals for lacking novelty despite reproducing similar data.
- The study explains how fine-tuning experimental parameters can introduce confirmation bias that produces apparent breakthroughs.
- Recommendations include releasing comprehensive datasets, disclosing full study volumes, and explicitly testing alternative explanations for reported signals.