Dear Rahat,

SEM is a very strict model. If you are variables are inconsistent, the fit of the model will always be poor. You can have a look on the modification indices, they may give you indications of how to improve your model. However, I would indicate two different ways:

1- If you would like to run a SEM, you could run a Bayesian approach, which will give you a satisfactory model.

2- You could also run a regression analysis or GLM. I guess in your case, this might be the best solution.

Best,

@Dario Cecilio

Dear, thanks for your reply Sir. But need hep from you.

Firstly, do you have any link of you tube video to run the Bayesian Approach ? It would be helpful.

Secondly, I think model is so poor for SEM. It couldn't show the modification indices

That'she why I think the basic reason may be data inconsistency.

Waiting for your valuableresponse dear Sir.

@Mercedes Orús-Lacort

Thanks for your valuable response ma'am.

Do you have any reference file for additional reading. Will really be helpful.

Thanks

Dear Rahat,

I'm not sure whether SPSS is running Bayesian SPSS. But you can have a look at Mplus and Mplus guide (link below). There you will find more information regarding Bayesian SEM. But if you are not getting even get the MI, perhaps you should just run a GLM model. It might be easier and give you a lot of information.

https://www.statmodel.com/download/usersguide/Mplus%20Users%20Guide%20v6.pdf

Best,

Thanks dear all

@ Dario Cecilio Fernandes & @ Mercedes Orús-Lacort