A team at the National Institute of Standards and Technology (NIST) has identified a pervasive statistical error in nanoscience data analysis — one that systematically weakens the apparent relationship between a nanoparticle's size and its performance — and has published mathematical corrections that reverse the effect.

The work appears in ACS Nano (published online Aug. 6, 2026; DOI: 10.1021/acsnano.5c12141), authored by A.L. Pintar, A.C. Madison, C.R. Copeland, N. Farkas and S.M. Stavis.

The problem sits at the first step of most nanomaterial studies. Researchers measure particle size, then plot a performance metric — brightness of a luminescent particle, drug payload, current through a nanoscale wire — against that size to extract a trend. The common mistake, the NIST team found, is assuming those size measurements are perfectly precise and accurate. In practice, measurements of objects a few to a few hundred nanometers across — the size range of many viruses — carry limited precision and accuracy, sometimes very limited. Treating noisy size data as exact biases the fitted trend toward zero, attenuating the apparent size dependence.

"These overlooked sizing errors have the effect of jostling a light bulb's dimmer switch," said electrical engineer Andrew Madison. "The errors dim the apparent relationship between the size of a particle and its brightness."

The practical consequences are concrete. "If a nanoparticle carries drugs for a medical treatment, for example, analysis of the dose could be mistaken," said physical scientist Samuel Stavis. "Or if a nanoscale wire was measured imperfectly, understanding of how much current flows through it could be off."

After noticing the flaw in a few key publications, the team surveyed several other studies covering various nanomaterials and found the same problem. The affected analysis underpins work across electronics, ceramics, paints, plastics, chemical catalysis and drug delivery — a global nanotech market the team cites at more than $100 billion annually and growing.

The correction

The team derived and tested mathematical corrections built on measurement error models. By quantifying the limited precision and accuracy of particle size measurements, the correction reverses the attenuation and reveals the true relationship between particle property and size.

Applying it requires an estimate of the measurement error. "Many studies begin by measuring the size of a nanomaterial," said mathematical statistician Adam Pintar. "But these measurements are imperfect, and sizing errors matter." The best route, Pintar said, is to obtain a reliable reference material for the particles and measure it with the same instruments and methods used in the study; the difference between those results and the certified reference values yields the error estimate. NIST sells a wide selection of reference materials, as do other organizations. Where no reference material exists for a particular nanomaterial, researchers can still apply the correction by carefully calculating the uncertainty of their sizing measurements.

Why it matters for the bench

The physics of the nanoscale drives the stakes. A material a few nanometers across often behaves differently from its bulk form — nanoscale gold, for instance, can appear ruby red rather than metallic gold. Exploiting that size-dependent behavior is the entire premise of nanotechnology, whether in quantum-dot color tuning or carrier dosing. If the statistical pipeline that extracts size dependence from experimental data systematically understates it, product development and material screening inherit the bias. The correction costs nothing but an error estimate — reference-material measurements already routine in calibrated labs — yet it changes the numbers that come out of the correlation analysis.

The NIST team frames the finding in metrological terms: a fundamental problem at the foundation of a research field, paired with a practical remedy. "It is a big ask to confront an overlooked problem at the foundation of a field of research," Stavis said. "But other fields of research have done so, and we offer a practical solution."

The development raises an adoption question for anyone publishing nanomaterial characterization: will journals and reviewers begin requiring measurement error models — or at least an uncertainty estimate on size measurements — as a condition for accepting size-property correlations? Labs that already own reference materials can answer that question cheaply.