IBM SPSS Statistics (as "SPSS" is now called) has no commands for SEM or Generalized SEM. SPSS users who want to do SEM often purchase AMOS, which is also owned by IBM. But AFAIK, AMOS still has no options for doing Generalized SEM.

SPSS users who want to estimate mediation models often turn to the macros provided by Andrew Hayes (e.g., PROCESS). But AFAIK, none of those macros are for OLS models, so will not be suitable for your logit model.

Finally, you might have better luck posting to a forum that is exclusively about SPSS. A convenient way to interact with the SPSSX-L forum is via the Nabble archive:

• http://spssx-discussion.1045642.n5.nabble.com/

HTH.

As there are no latent variables, I'm not sure the PROCESS approach is necessary. It gets you a CI for the indirect effect, but you can all the path coefficients straight from SPSS logistic regression. A significance test of the indirect effect is simply the joint test of significance for the ab path.

i.e., a x b = indirect effect and if the a path and b path are both p < .05 then the indirect is significant (with alpha = .05; as if either path is NS the overall indirect effect is logically NS).

Also as you have no latent variables then you could do this with piecewise SEM. This would be pretty easy to set up in R. See

https://jslefche.github.io/sem_book/index.html

I'm not sure if there is an easy way to do it in SPSS.

Thom Baguley, I have never used PROCESS, but it deals with manifest variables only, AFAIK. Here is an excerpt from the main webpage for it, for example:

PROCESS is an observed variable OLS and logistic regression path analysis modeling tool. It is widely used through the social, business, and health sciences for estimating direct and indirect effects in single and multiple mediator models (parallel and serial) , two and three way interactions in moderation models along with simple slopes and regions of significance for probing interactions, and conditional indirect effects in moderated mediation models with a single or multiple mediators or moderators.  The use of PROCESS is described and documented in Introduction to Mediation, Moderation, and Conditional Process Analysis, published by The Guilford Press.  PROCESS was written by Andrew F. Hayes.  Source: https://www.processmacro.org/index.html

I did not know that PROCESS now includes logistic regression (first line above), so it may in fact give Md Jahirul Islam what he is seeking.