Dressing-up disinformation: the contextual presentation of lies

This paper focuses on a historically documented tactic that professional deceivers rely on: presenting inaccurate claims together with accurate claims. While historical sources often point to this as a textbook-method to make inaccurate claims look more believable, there has been relatively little experimental research on its effectiveness and the background psychological processes. In three pre-registered online experiments (N=817), we found evidence to the existence of the dressing-up effect. Controlling for a variety of alternative explanations, we also found that it is not sensitive to the length of the message, or the order in which the claims are presented, but disappear once the accurate dressing claims and the inaccurate target claim are uttered by different sources. Furthermore, the effect persisted even when the participants’ accuracy motives were stimulated using a monetary reward. These findings are compatible with a rational, Bayesian model of belief updating on the basis of testimonies, following which belief are updated in view of the credibility of the communicator. We entertain the idea that the direction of this interaction between source and content is guided by a rule of least resistance: in any one time, humans aim to update their beliefs as little as possible.