Partial Correlation vs. Multiple Regression

What is the purpose of doing JUST a partial correlation as opposed to a full multiple regression? (In cases where there is a clear outcome variable.)

It seems to me that doing a multiple regression is the most appropriate, because the two measures are equivalent, but working through the full regression analysis allows one to ensure that all assumptions have been met (e.g., homoscedasticity).

Yet I seem to recall seeing (Psychology) articles in which the partial correlation is reported by itself.

So, what am I missing? Why do just a partial correlation?
I think it depends on what you are trying to achieve. Are you trying to find causation, do predictions etc ? If so then regression is more appropriate.
Thanks for your answer. Why is regression more appropriate for finding causation and making predictions?

I'm not necessarily looking for causation, but I still think about the safety of going through the assumptions first. However, if partial correlations are safe and just as good, they are certainly easier to do.