I was running mediation models and no surprise that coefficoients dfiffer... and majority of significance values are the same....
However, when I looked at indirect effects significance value slightly differ.... slightly? well, 0.47 and 0.53 is a crucual difference in social sciences....
Could anyone provide some explanation?
Thank you
Could you give more detail, and explain how you define your random variables observed vs your "standardised" random variables. Do you assume Gaussian processes, and divide (x-m) by sigma, or something similar? Intuitively I would not normalise or standardise, but keep actual amplitude observed. However, this depands on the detail of your question.
check data is normalized or not? If data is not normalized this kind of error occurs.
Maybe because the date is not normalized. But do you run CB-SEM (MPLus keyword)... then data should have been normalized.
Could you give more detail, please.
- Badges
- Science method
- Similar topics
- Mathematics
- Statistics
- Statistical Techniques
- Structural Equation Modeling