Validity

Research

Whether a study or measure actually captures what it claims to, and whether conclusions hold up (construct, internal, external validity).

Sources

1. Stanford Encyclopedia of Philosophy — “Measurement in Science”, Eran Tal (first published 15 June 2015; substantive revision 7 August 2020) plato.stanford.edu accessed 22 September 2026

For the measurement sense of validity, Tal’s entry gives Denny Borsboom’s realist formulation: “a test is valid for measuring an attribute if and only if a) the attribute exists, and b) variations in the attribute causally produce variations in the outcomes of the measurement procedure.” That is a demanding standard — it makes validity a causal claim about the world, not a property of the test’s statistics. Only the measurement sense is sourced here. The internal and external validity of study designs, also named in this entry, are not covered by this source.

The Stanford Encyclopedia of Philosophy is cited because it carries a named author and a dated revision for each entry, so the claim is attributable to a person on a date — which is the test this lexicon applies to reference sources, and the reason an open-edit encyclopedia is not used. The application to research practice above is this lexicon’s reading.

Related terms

• Abbreviation
• Abstract
• Abstraction
• Academic Database
• Analysis
• Annotation

Also defined as

The quality of being logically sound, or of accurately measuring what a study claims to measure (internal, external, construct validity).

Also defined as

In logic, whether a conclusion follows from its premises; in research, internal, external, and construct validity — and reliability is not validity, since a measure can be perfectly consistent and still measure the wrong thing.

From the AJG lexicon archive (July 2026).