I read through the study about which this article reports. The first two experiments are not very robust. They present a _plausible_ association between cursi

I read through the study about which this article reports. The first two experiments are not very robust. They present a _plausible_ association between cursing and honesty, but it's not especially convincing. There is so much potential for confounding variables and sampling defects (e.g. self-reported data from Mechanical Turk users, a group who may not represent the population at large) that you could drive a truck through it. There are likely many other plausible explanations for the results that don't actually depend on an association between cursing and honesty. For instance, in the first experiment, participants were asked question such as "Are all your habits good and desirable ones?" If they answered, "Yes," it was considered a lie. But it might alternatively indicate several other things, such as poor reading comprehension, paranoia about how the answers might be used, poor self-inventory skills, etc. As for the third state-level experiment, it's just silly. It's the sort of thing you might see on Buzzfeed or my own site, Correlated.org. Don't take it seriously.