It is not required, but highly recommended to present results on
simulated data, even if you are analyzing "real" data. The reason is
that simulated data give you an opportunity to illustrate and confirm
properties of a method. If you can't understand and predict the
behavior of your method on simulated data, there is little hope of
doing so on other data sets.
-John