Where do the numbers on this site come from?
An evidence index is a list of the primary sources a set of claims rests on, each quoted in its own words, dated, and paired with a statement of what it does not say. It lets a reader check an argument against its sources instead of against the confidence of whoever is making it.
Hiring-assessment marketing runs on a small number of studies and regulations, quoted second and third hand, usually without the sentence that limits them. This page lists the ones this site argues from. Each entry gives the full citation, the source's own words with a locator, what it is commonly claimed to say and does not, whether you can read it without paying, and which pages here depend on it.
Every quote below was read at the primary source on the date shown beside it. Where the full text sits behind a publisher, that is stated rather than glossed over — a citation you cannot reach is a citation you cannot check.
Founder of AssessAll and of Bodhih Training Solutions, a corporate training company in Bangalore. Works on assessment design, scoring and reporting across hiring, L&D and certification programmes.
Last reviewed
Meta-analysis
Sackett et al. (2022) — the validity re-estimates
The paper that cut the published validity of most hiring predictors by .10–.20 and put structured interviews on top.
Schmidt & Hunter (1998) — the table everyone still quotes
The 1998 meta-analysis behind almost every validity figure in assessment marketing — and the reason quoting it in 2026 is a dated claim.
Journal commentary
Federal regulation
Statute
6 questions to ask of any number a vendor quotes you
These work on anyone's citation, ours included. Put them to the next validity coefficient that appears in a sales deck, including the ones on this site.
- 1
Does the citation name a source you could actually open?
A number attributed to 'research shows' or to a consultancy blog that cites another blog is not a citation. Ask for the author, year and journal or the regulation's section number.
- 2
Is the quoted sentence in the source, or is it someone's summary of it?
Summaries drift. The most reliable check is to find the sentence in the original and see whether it carries the qualifications the summary dropped.
- 3
How old is the estimate, and has the field revised it since?
Validity coefficients in personnel selection were revised downward across the board in 2022. A figure that was standard for twenty years can be superseded without the market noticing for several more.
- 4
Is it a corrected figure or an observed one?
Most published validity coefficients are corrected for range restriction and criterion unreliability. The correlation you would see in your own pipeline is lower, and the size of the gap depends on assumptions worth asking about.
- 5
Does the source say the thing is generally regarded, or that it is?
Regulations in particular are written in hedged language for a reason. A guideline saying a ratio 'will generally be regarded as evidence' is not a guideline saying the procedure is unlawful, and the difference is the whole meaning of the paragraph.
- 6
What does the source say that the person quoting it left out?
The omitted sentence is usually the one that limits the claim. In the four-fifths rule it is the sentence saying smaller differences may still count; in the 1998 validity table it is the column showing a different instrument adding more.
Where to go next
- The measurement glossary — what the terms in these sources mean, defined one at a time.
- The free calculators — run the arithmetic in these papers on your own numbers, in your browser.
- Annotated sample reports — what happens to these numbers once they reach a score report.