The theoretical difference is somehow clear for me, but unfortunately, I've not still been able to figure out what's the difference between these two methods in terms of their application in statistical packages.

Let me explain it using my research example, which consists of seven latent variables, and each one contains a considerable number of items/indicators as usual of socio-psychological phenomena. The path model could not be run using indicators and their latent constructs in Lisrel, but it could be run when I create a composite variable out of the indicators for each individual latent variables in SPSS.

- Is this latter method Path Analysis? In other words, the practical difference between SEM and Path Analysis is this fact that in case of a Path Analysis we have to compute a composite variable for latent variables and in case of SEM we must not?

- The other question of mine is whether composite variable should be computed using weighted or unweighted mean?

Thank you in advance,

Similar questions and discussions