Yes

Sipahi Dr. Esra Thank you for your reply but PROCESS macro has already provide the answer for regression in the analysis right ? So may I know why I need to run a separate analysis ?

Thank you :)

No, there's no strict rule mandating that you run linear regression before doing PROCESS or any kind of mediation since mediation results will provide you with separate components, indirect effects, and direct effects which state the same information as linear regression models (albeit probably with slightly different coefficients and p-values due to differences in calculation). However, a lot of people do run and show the results of their regression models because it just shows separate regression models for each path in your mediation model.

It's entirely up to you to decide if you want to pair your mediation analyses with path-by-path linear regression models.