What is Psychometrics?
Psychometrics is the science of measuring psychological attributes — ability, personality, knowledge, attitude — that cannot be observed directly. It covers how questions are written and calibrated, how responses become scores, how much error surrounds a score, and what evidence is required before a score may support a decision about a person.
The two halves of the field
One half is about instruments: item writing, scaling, calibration, reliability, dimensionality, fairness analysis. The other half is about inference: what a score licenses you to conclude, and for which use.
Most commercial disputes about assessment are really disputes about the second half. An instrument can be technically sound and still be used to support a conclusion it was never validated for, which is why validity is defined in relation to a purpose rather than as a property of the test.
What a psychometrically literate buyer asks
What is the construct, stated in one sentence. What reliability figure is claimed, of which kind, on how many items and how many people. What the norm group is. What the standard error of measurement is in the units of the reported scale. What validity evidence exists for this use, not for the instrument in general.
Five questions, all answerable in a sentence each by a vendor who has the evidence, and a documented refusal by one who does not.
There is an international standard, and half of it is addressed to the buyer
ISO 10667 — “Assessment service delivery — Procedures and methods to assess people in work and organizational settings” — is published in two parts, and the unusual thing about it is which way round they point. Part 2 sets requirements for the service provider. Part 1 sets requirements for the client: ISO 10667-1:2020 “establishes requirements and guidance for clients working with one or more service provider(s) to carry out the assessment of an individual, a group or an organization for work-related purposes.”
That is a standard saying the buyer has obligations of their own — about stating the assessment need, the conditions under which it will be delivered, the competence required of the people interpreting it, and what happens to the data afterwards. Almost no vendor material mentions it, for the obvious reason that it is not a badge anyone can display.
It is worth knowing about for two practical purposes. First, it gives a procurement team a published frame for questions that otherwise sound improvised. Second, when a vendor claims alignment with it, you can ask which part and which edition — Part 1 reached its second edition in November 2020, and the 2011 edition of Part 2 was withdrawn in 2020, so a citation to the 2011 text is a sign of material that has not been refreshed in five years.
What the field calls itself, and why the label is contested
“Psychometrics” names the measurement discipline: item calibration, scaling, error estimation, validity argument. It is routinely used in commercial material as a synonym for “personality questionnaire”, which is a category error — a numerical reasoning test, a knowledge test and a structured-interview rating scale are all psychometric instruments, and a four-quadrant workstyle profile may be a useful conversation tool without carrying much psychometric evidence at all.
The distinction matters when reading a claim. “Our assessments are psychometric” describes a method of construction, and says nothing about how well any particular instrument was constructed. The five questions above are what convert the adjective into something checkable.
Sources
Read next
Related terms
More terms beginning with P
Check a selection process against the four-fifths rule
Free, no signup, computed in your browser — with the remedy, not just the verdict.
Last reviewed