Endogenous variables in causal statistical modeling are variables that are hypothesized to have one or more variables at least partially explaining them. Commonly referred to in econometrics and the structural equation modeling family of statistical techniques, endogenous variables may be effect variables that precede other endogenous variables; thus, although some consider endogenous variables to be dependent, such a definition is technically incorrect.

Theoretical considerations must be taken into account when determining whether a variable is endogenous. Endogeneity is a property of the model, not the variable, and will differ among models. For example, if one were to model the effect of income on adoption of environmental behaviors, the behaviors would be endogenous, and income would be exogenous. Another model may consider the effect of education on ...

  • Loading...
locked icon

Sign in to access this content

Get a 30 day FREE TRIAL

  • Watch videos from a variety of sources bringing classroom topics to life
  • Read modern, diverse business cases
  • Explore hundreds of books and reference titles