A latent variable is an unobserved construct inferred from a set of observed measurements through statistical modeling. It represents underlying properties or dimensions — such as intelligence, socioeconomic status, or psychological disposition — that cannot be directly measured but are estimated from patterns in observable indicators. Parameters: (1) latent construct not directly measurable; (2) identified through covariance structure among observed variables; (3) estimated via factor analysis, structural equation modeling, or item response theory. Persistence: mathematical formalism encoded in statistical software and peer-reviewed methodology, transmitted through academic disciplines and training programs.[formal: latent_variabilis | substrate: mind | horizon: a life | explicit: yes | epoch: 0.01]
Accepted ontology entry
latent-variable
A latent variable is an unobserved construct inferred from a set of observed measurements through statistical modeling. It represents underlying properties or dimensions — such as intelligence, socioeconomic status, or psychological dispos…
Definition
Why it is in scope
A human-made statistical concept for an unobserved variable inferred from observed measurements through mathematical modeling. Used to capture constructs that cannot be directly measured — such as intelligence, socioeconomic status, or psychological traits — enabling quantitative analysis of abstract phenomena.
Names and aliases
- latent-variableen · CANONICAL
Relations from this entry
- cmsnqj43w04nr1q13drcajl3kDEPENDS_ON →
Latent variables are unobserved constructs inferred through measurement models. Remove measurement-model — latent variables have no operationalization mechanism and stop operating as usable concepts. Per Law 8 removal test.
Relations to this entry
- cmsnoay9h04ha1q13xkt96h93← DEPENDS_ON
SEM operates by specifying and estimating relationships among latent constructs. Remove the concept of latent variables and SEM ceases to operate — it has no target constructs for its measurement and structural models. This passes the removal test (Law 8): not merely sayability loss but operational collapse.
- cmsnqj43w04nr1q13drcajl3k← DEPENDS_ON
Measurement model operates by specifying the relationship between latent constructs and their observed indicators. Remove the concept of latent variables and the measurement model ceases to operate — it has no constructs to measure. The removal test (Law 8) passes.
Record identity
- Created
- Aug 10, 2026, 8:55 PM UTC
- Content hash
- 0612d13ea512d3951aa24862ae1d8ff6c584b05fafa69b608603236704195703