@Helen Dunne You can use any of multiple regression and SEM for testing your Hypotheses. But SEM has some advantages over multiple regression as it can test a complex model with two or more dependent variables at the same time while multiple regression in SPSS won't do this. Moreover, for mediation analysis in SPSS, you'll be in need of PROCESS Macro and you can test only one mediation model at a time, whereas in SEM (Using AMOS or any other software) you'll be able to test multiple mediation models along with direct effects of predicting variables. Furthermore, AMOS also accounts for error terms and SPSS doesn't hence the results from SEM are more robust than the results of multiple regression.

It all depends on your hypothesis objectives.

It depends on the number of variables and hypothesis you are testing. For complex problems, Structural Equation Modeling (SEM) is preferable over Multiple Regression analysis.

Dear Helen

In my humble opinion, it is recommended to use multiple regression as all your variables are observed variable.

SEM is usually used for latent variables.

You may want to have a look at this paper:

Jin, S., Noh, M., Yang-Wallentin, F., & Lee, Y. (2021). Robust nonlinear structural equation modeling with the interaction between exogenous and endogenous latent variables. Structural Equation Modeling: A Multidisciplinary Journal, 1-10.

Hope it helps.

All the best.

Regards

Mei Peng

I think using SEM, especially (PLS) better for your hypothesis because you have mediation in your model, moreover having latent variable or not, doesn't the only reason to use SPSS or SEM, but SEM has some features easy to handle of the latent variables.