I would like to ask you if it is possible to use standardised z scores in MPLUS instead of raw scores for variables included in a mediation model.

I work on a study, where we have measured stress in 3 measurement points in an exam situation (pre-stress – 5 days before the exam, anticipatory stress – on the exam day before the exam and poststress – on the exam day after the exam), we have constructed residualized stress change scores with simple linear regression and we have computed stress reactivity (as the residual of the anticipatory stress predicted by the prestress score) and stress recovery (as the residual of the poststress predicted by the prestress score). In the second step we have tested a mediation model in MPLUS with standardized residuals for stress recovery as the dependent variable and standardized z scores (transforming the raw scores to z scores) for the independent and mediator variables.

My question is if it is possible to test a SEM mediation model in MPLUS with standardised scores. Is there a difference in reporting the results if standardized scores have been used?

Thank you.


Yes, it's possible.

You will lose information about variances, which you might care about. You should not do this if you have a multiple group model.

If you use standardized variables, you can obviously only report standardized estimates.

| cite | improve this answer | |

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.