How the Chavery Score is calculated

Version 1.0. Six components, weighted, each on a nought-to-ten scale. Nothing is generated and nothing is judged: the same figures always give the same score, and the whole formula is on this page so you can check the number rather than take it.

The components

Each is scored with the same curve — 10 × v ÷ (v + h) — where h is the value that scores exactly five. That single parameter is the whole of each component’s opinion, which is why it is published: you can disagree that twelve citations per work is the midpoint, and that is a conversation about a number rather than about a black box.

Career stage, not seniority

A score built on accumulated citations measures how long somebody has been publishing. So citation impact is citations per work, not the total, and sustained impact is the m-quotient — h-index divided by years publishing, which is Hirsch’s own answer to the same problem. Together they carry half the score. The effect is that an active mid-career researcher can outrank a senior one with three times the works and six times the citations, which is the intended behaviour rather than a side effect.

• student — under 4 years since the first publication
• early career — under 8 years since the first publication
• mid career — under 16 years since the first publication
• senior — everything beyond that

When there is no score

A researcher whose citation count Chavery does not hold has no score, not a low one. The two required components are Citation impact and Sustained impact, and without them the score is withheld and the reason is reported in its place. Scoring an unknown as a zero would print a measurement of Chavery’s ignorance as a measurement of a researcher, and nothing about the output would look wrong.

A genuine zero is still zero. Somebody with published work and no citations yet is measurable, and low.

What version 1.0 does not do

It is not normalised for field. Citation rates differ several-fold between, say, molecular biology and mathematics, and correcting for that needs per-field baselines Chavery does not yet hold. Rather than approximate it and imply a precision that is not there, v1.0 states the limitation — and every score carries the version that had it, so a later formula can fix this without rewriting anybody’s history.

It also does not measure teaching, supervision, clinical work, peer review or anything else a researcher does that is not a published, indexed output. A Chavery Score is a narrow thing measured carefully, not a verdict on a career.

A worked example

A mid-career researcher: 28 works over 13 years, 312 citations, h-index 9, 17 works in the last five years, first or corresponding author on 12, two verified identifiers and one self-reported.

Chavery Score 5.49 / 10 — mid career, 13 years publishing. Computed by the same code that scores every profile; this page is generated from it, so the two cannot disagree.

Where the figures come from

Publications and authorship come from works claimed on a Chavery profile, imported from Crossref by DOI or from a researcher’s ORCID record. Citations, h-index and i10-index come from OpenAlex, matched by a verified ORCID and never by name — name matching would eventually attach one researcher’s citations to another with a similar name.

Nothing here is self-reported. A researcher cannot type in their own metrics, and a Google Scholar profile is stored as a link rather than read, because Scholar has no public interface that would let Chavery confirm the figures or the ownership.