I am an elementary education phd candidate. I am sorry for my basic question but i am pretty unfamiliar to multivariate statistics.
I was trying to find a answer to my question over years. As i stated above, i am not well on advanced statistics. I hope you can help me.
I have applied a new learning method in elementary school. I have 4 groups, 3 of them are experimental. This is quasi experimental pre post test design. I have applied a new learning method in education and tested in both pre and post. However i will only use 2 of them to validate model. Here is my model which drawn by a famous researcher in my research area. https://ibb.co/S76V46M
Group goals etc. is independent variable which i am currently testing effectiveness of that. Social cohesion, motivation and peer relationship parts are dependent and latent variable which effects directly pupils' academic achievement. Enhanced learning refers to academic achievement (observed variable) which is measured with a test. All measurement tools are applied both in pre and post phase.
I have learned that Mplus could help me to validate this model via growth modeling time series model (Maybe i am wrong, sorry).
Is it possible to validate this model via Mplus? I have not purchased software yet, if it is possible, is there any example syntax for help?
Growth modeling would not be relevant with only 2 time points. But the modeling seems doable in a general structural equation framework for the 2 time points. For each time point it looks like a "mediation" model. See our Topic 1 Short Course video and handout on our website. for this general analysis strategy question, you may also want to ask on SEMNET. When you do, you need to explain what the double-headed arrow means.