in a model having three independent variables and one dependent variables there is a mediator so i want to measure its impact through SPSS analysis.kindly explanation required which is the best method to get optimum results?
thanks dear much appreciated Mohammed Abdulrab.i try to find the book mentioned in the link as add the macro.and i have added the Andrew Hayes process.
SEM(AMOS or PLS) use to analyse the casual relationship between the variables as a model, it allows to analys latent constructs, and SEM used when the model is complex.
It is much better when you use AMOS or smart_pls but there are justifications for using which one is suitable that based on some issues such as if your study aims to testing theory that leads to Using AMOS IF NOT , USE PLS and also normality distribution...etc
With these measures a fit value can be obtained for a saturated model. Then one can compute the measure for each of the following models: no direct effect, no effect from causal variable to the mediator, and no effect from the mediator to outcome. Then the best fitting model is the one with lowest value.