This paper introduces a new concept of full implementation that takes into account agents' preferences for understanding how the process works. We assume that the agents have intrinsic preferences for honesty in the sense that they dislike the idea of lying when it does not influence their welfare but instead goes against the intention of the central planner. We show that the presence of such preferences functions effectively in eliminating unwanted equilibria from the practical perspective, even if the degree of the preference for honesty is small. The mechanisms designed are detail-free and involve only small fines.