作者: Dennis J. Aigner , Cheng Hsiao , Arie Kapteyn , Tom Wansbeek
DOI: 10.1016/S1573-4412(84)02015-8
关键词:
摘要: Publisher Summary This chapter discusses latent variable models in econometrics. The essential characteristic of a is revealed by the fact that system linear structural equations which it appears cannot be manipulated so as to express function measured variables only. It for equation called “latent model,” there at least one more independent than number variables. Usage term “independent” contrasted with “exogenous” variable, common phrase econometrics, includes measurement errors and residuals themselves. In functional model, true values exogenous are fixed variates, therefore, best thought nuisance parameters may have estimated en route getting consistent estimates primary interest. Finally, restrictions on model's covariance structure, commonplace sociometric psychometric modeling, also serve aid identification.