HI everyone. Currently I am doing a mediation research paper using SEM for my Master Degree. I notice that some papers doing mediation analysis with SEM report the model fit why some others don't. Some paper even use item parceling ( which, to be honest, I don't understand this technique very much) to achieve a model fit for their models.

When looking more deeply, I found out that Kenny said "Normally with SEM, one computes a measure of fit. However, the mediational model is saturated and a no usual measure of fit is possible. However, one can adopt the following strategy. Use a fit statistic an "Information" measure like the AIC or BIC. " So it seems that one doesn't need to report the model fit to publish their paper.

So what I want to ask is whether it is necessary to achieve the model fit for your mediation analysis when using SEM. And if your model is a simple model (with 2 IVs, 1 M and 1 DV), should I use item parceling in order to get the fit indices.