Comment by mseebach

Yes, of course, if an experiment gets the wrong result, you declare the experiment faulty and try again. You should never let your ideas become polluted by empirical evidence.