This is the first time I will be using SEM using AMOS for my research.
My research questions are:
Hypothesis 1: There is a positive relationship between the materialistic value-orientation and perceived social stress.
Hypothesis 2: Religion mediates the effects of materialism on perceived social stress.
Hypothesis 3: The mediating effect of internal locus of control on the association between materialism and perceived social stress is higher among those who possess weaker religiosity in comparison to those who have stronger religiosity.
So Independent variable (materialism), Dependent variable (Stress), Mediator (religion), and then Locus of control has mediating effect on strong religiosity and weak religiosity on Stress.
I have conducted the survey and each Variable have 7-15 items each as well as age, gender, education level and religion (given the research questions).
Now my question is from experienced people with SEM and AMOS:
1) How should I model this construct in AMOS to explore this hypothesis? I am particularly confused on how to model hypothesis 3. 1 and 2 is pretty straight forward (Materialism to religiosity and Stress level and materialism to stress level paths)
2) Other than conducting CFA, and regression and goodness for fit test; what other tests would provide fruitful information in AMOS?
3) For the Variables should I input all the items from the survey or only the ones most important?
Thank you for any insight.