Experimental research give you a feeling of some kind of accomplishment. Eventually you will have a feeling that really you have physically done something good.
From my experience in numerical research it's difficult to published because it's necessary the validation of numerical results. However for the experimental it's more easy to publish the results of it's new.