Science

Current Picks

Twenty-Seven Crystal Growth Runs Traced One Superconductivity Reproducibility Gap

By Alice Chen / Jul 18, 2026

How a new instrument with real-time oxygen monitoring traced a reproducibility gap in superconducting crystal growth, and what it means for funding structures.

A Budget Office’s Sieve Standard Split Two Ocean Sediment Core Chronologies

By Karim Osman / Jul 18, 2026

How a budget office's tolerance threshold for sediment core age models exposed hidden assumptions in paleoclimate dating, sparking debate over whether the field's consensus standard filtered out legitimate variability.

A Single Photometric Calibration Star Divided Two Exoplanet Atmosphere Spectra

By Karim Osman / Jul 18, 2026

How one calibration star enabled two exoplanet atmosphere spectra, revealing water, sodium, and haze. A methodology piece on the hidden role of stellar templates.

A Flat Grant Overhead Rate Merged Two fMRI Anxiety Protocols

By Karim Osman / Jul 18, 2026

How a flat institutional overhead rate forced two anxiety fMRI studies into one hybrid protocol, diluting specificity and raising questions about funding incentives in neuroscience.

A Single Lab Ventilation Glitch Altered Two Mouse Learning Curves

By Renu Shah / Jul 18, 2026

A lab HVAC failure skewed mouse memory data, revealing how cage microclimate masks genotype effects and threatens reproducibility in behavioral neuroscience.

A Single Budget Overhead Cap Splits Two Climate Code Reproducibility Analyses

By Renu Shah / Jul 18, 2026

How a 15% NSF overhead cap on subawards forced two climate modeling teams down diverging paths—one verified, one retracted—over a difference of roughly $8,000 in compute funding.

A Grant Reviewer's Catalyst Purity Clause Scuttled Two Synthesis Labs

By Alice Chen / Jul 18, 2026

How a single clause demanding trace metal purity in catalysts derailed two synthesis labs, costing 18 months of work and sparking debate over funding gatekeeping.

Twenty Psych Lab Equipment Budgets Forced One Replication Protocol Off a Second Registry

By Karim Osman / Jul 18, 2026

How equipment costs forced a replication protocol off PsychFileDrawer. The Many Eyes project aimed to replicate 20 studies; only 14 finished. Budgets, not theory, were the bottleneck.

Five Kaleidoscope Phase Plates Reveal One Quantum Optics Measurement Angle

By Alice Chen / Jul 18, 2026

A controversial quantum optics method uses five kaleidoscope phase plates to measure a single angle. Replication attempts reveal hidden parameters and procedural choices that split the field.

One Desk Drawer Holding Two Magnetometer Calibration Constants

By Karim Osman / Jul 18, 2026

A 0.7% offset between two calibration constants for the same magnetometer sat unnoticed for years. How funding gaps and career incentives buried a discrepancy that could reshape exoplanet magnetic field estimates.

Sixteen Bat Night-Roost Surveys Funded One Statistical Power Calculation

By Karim Osman / Jul 18, 2026

A single power calculation cost $4,800, funded by sixteen bat roost surveys. This article examines the hidden trade-offs in ecology research funding between data collection and statistical inference.

Ten Amphipod Species Disappeared from One Corrected Sediment Core Chronology

By Renu Shah / Jul 18, 2026

A revised chronology of a Lake Greifensee sediment core reveals ten amphipod species vanished abruptly, not gradually. The finding underscores how dating precision can flip paleoclimate narratives.

One Climate Code Branch Forced Two Ocean Models Onto Different Turbulence Closures

By Alice Chen / Jul 18, 2026

A code fork around 2005 split two major ocean models onto different turbulence closure schemes. The choice of closure amplifies over decades, affecting hindcasts and reproducibility.

A Single fMRI Slice-Timing Parameter Split Two Lab Anxiety Studies

By Karim Osman / Jul 18, 2026

Two labs studied anxiety with fMRI and got opposite results. The only difference was a slice-timing correction parameter in SPM. A replication audit reveals how a tiny default split the field.

A Single Tide Gauge Rental Fee Shifted Two Sea Level Acceleration Curves

By Alice Chen / Jul 18, 2026

How a $15,000 annual rental fee for two tide gauges, lost when a grant renewal failed, introduced a hidden discontinuity in Pacific sea level acceleration curves—and what it reveals about the fragile economics of long-term climate observation.

A Gut Microbe Enzyme Rate Shifts Two Lab Mouse Anxiety Assays

By Alice Chen / Jul 18, 2026

A single gut bacterial enzyme produces contradictory results in two standard mouse anxiety tests, raising questions about how we measure anxiety-like behavior in rodents.

Seven-Line Vignette Bias Shrank One Behavioral Economics Replication

By Alice Chen / Jul 18, 2026

A seven-line vignette in a 2012 behavioral economics study may have contributed to its failure to replicate. Weak treatments, small samples, and flexibility in analysis all played a role.

One Sieve Mesh Size Reassigned Two Hundred Polymer Viscosity Measurements

By Alice Chen / Jul 18, 2026

A single sieve mesh size change in a polymer lab reassigned 200 viscosity measurements. How procedural choices in materials science produce data that can shift 15–30%.

A Bureau of Land Management Drilling Fee Split Two Seismic Hazard Forecasts

By Karim Osman / Jul 18, 2026

A BLM drilling fee in Oklahoma triggered a split between two seismic hazard models—one from USGS, one industry-funded. The disagreement delayed a permit and exposed deeper tensions in earthquake risk assessment.

A Sieve Mesh Size Swapped Two Ocean Circulation Records

By Renu Shah / Jul 18, 2026

Two studies analyzing the same sediment core reached opposite conclusions about Atlantic Ocean circulation. The culprit: a 63 µm versus 150 µm sieve mesh that selected different foraminifera size fractions.