Examples
Worked examples
- Is an instance
A drug efficacy effect observed in US trial populations and confirmed in Sub-Saharan African populations.
- Is an instance
A behavioural finding from undergraduates replicated in older non-Western samples.
Counter-examples
Looks similar, but isn't
- Not an instance
Re-running the analysis (reproducibility).
- Not an instance
Re-collecting data with the same demographic profile (replicability, not generalisability).
Editorial commentary
Generalisability — also called external validity or, in the causal-inference literature, transportability — is the extent to which a study’s findings hold true beyond the specific population, setting, and time period that were actually studied. A result can be highly robust and even successfully replicated within the exact conditions it was first observed in, and still fail to generalise: an effect measured in undergraduate participants in one country may not appear in older, non-Western, or clinical populations; a drug’s efficacy in a controlled trial population may differ in broader real-world use.
Generalisability is a separate axis from the other reproducibility-family concepts it is routinely confused with. It is not the same as re-running the original analysis code on the original data (that is reproducibility), and it is not the same as re-collecting new data under the same population and conditions to see if the effect recurs (that is replicability). A study can be perfectly reproducible and even replicable, and still be narrow: true only for the specific sample it was run on.
Why it has become a live research-integrity question
Generalisability has moved from a methods footnote to an active concern partly because of what has been called a “generalizability crisis” in some fields: a large share of behavioural and social-science research draws on convenience samples (frequently university students, or participants recruited from a narrow set of countries) while presenting findings as if they describe people in general. Funders and journals increasingly expect authors to state the population and setting a finding is claimed to apply to, rather than implying unrestricted generalisability by default, and some registries and reporting checklists now ask explicitly whether a claimed effect has been tested in more than one population before it is described in general terms. Crowdsourced replications run across many countries and samples are one of the more direct ways a field can gather actual evidence about generalisability rather than assuming it.
References
- Shadish, Cook, Campbell, 'Experimental and Quasi-Experimental Designs for Generalized Causal Inference' (Houghton Mifflin, 2002).
- Yarkoni, 'The generalizability crisis' (Behavioral and Brain Sciences, 2022).
Also known as
external validity · transportability
Machine-readable encodings
Use in your systems
<role vocab="credit"
vocab-identifier="https://casrai.org/dictionary/"
vocab-term="Generalisability"
vocab-term-identifier="https://casrai.org/dictionary/term/generalisability" />{
"@context": "https://schema.org",
"@type": "DefinedTerm",
"@id": "https://casrai.org/dictionary/term/generalisability",
"name": "Generalisability",
"identifier": "https://casrai.org/dictionary/term/generalisability",
"description": "The extent to which a study's findings extend to populations, settings, or conditions other than those directly sampled.",
"inDefinedTermSet": "https://casrai.org/dictionary/domain/reproducibility#set",
"url": "https://casrai.org/dictionary/term/generalisability",
"sameAs": [
"external validity",
"transportability"
],
"license": "https://creativecommons.org/licenses/by/4.0/",
"publisher": {
"@id": "https://casrai.org/#organization"
},
"author": {
"@id": "https://casrai.org/#editorial-team"
},
"datePublished": "2026-05-21T02:22:49",
"dateModified": "2026-08-22T15:41:33",
"inLanguage": "en-GB",
"isAccessibleForFree": true
}
