Dismal Research

You usually read a disclaimer like "... if the true model is... bla, bla, bla...". This implies that all estimating models are wrong. You report the least wrong one. As simple as that. As dismal as it can be.