While the sample size is still small and thus conventional standards of statistical significance have not been reached, early indications are that, if one defines "effectiveness" as the indication that one has successfully made one's point, then saying "If you hear any Harry Potter spoilers and tell me, I'm going to be very annoyed" is less effective than "If you hear any Harry Potter spoilers and tell me, I will cut you."
BTW, I'm getting the book at midnight at Harvard Square. I'm thinking about doing an encoded simulblog as I read it.
Comments off for your protection!
