Rc. Maccallum et Mw. Browne, THE USE OF CAUSAL INDICATORS IN COVARIANCE STRUCTURE MODELS - SOME PRACTICAL ISSUES, Psychological bulletin, 114(3), 1993, pp. 533-541
In conventional representations of covariance structure models, indica
tors are defined as linear functions of latent variables, plus error.
In an alternative representation, constructs can be defined as linear
functions of their indicators, called causal indicators, plus an error
term. Such constructs are not latent variables but composite variable
s. and they have no indicators in the conventional sense. The presence
of composite variables in a model can, in some situations, result in
problems with identification of model parameters. Also, the use of cau
sal indicators can produce models that imply zero correlation among ma
ny measured variables, a problem resolved only by the inclusion of a p
otentially large number of additional parameters. These phenomena are
demonstrated with an example, and general principles underlying them a
re discussed. Remedies are described so as to allow for the evaluation
of models that contain causal indicators.