I don't like the explanation of why p-hacking is bad. p-hacking can be used for hypothesis formation, but it cannot be used as the test of the same hypothesis that it gave birth to. For that you need a separate data set.
The replication crisis is arguably centered around peoples' inability to grasp that very concept. You're pretty good at communication, though, that's one of the best ways I've ever heard it phrased. What's this guy got, like, zero grad students? Is that even possible for someone remotely well-established? No, no, I don't think that's the problem. Guy's just a fucker. It happens.p-hacking can be used for hypothesis formation, but it cannot be used as the test of the same hypothesis that it gave birth to. For that you need a separate data set.
They contacted Wansink and asked him to share his data, but he responded that it wouldn’t be possible because the data contained identifying characteristics of the research participants.
More or less. Sometimes that's OK, but typically not when people are involved. Especially when "measurables" resemble "now tell me, Karen, how good do you feel about this economic transaction on a scale of 1 to 2?". I'm actually finding it difficult to make more sweeping statements, here. Experimentation sucks like that. How about: Every experiment should be uniquely designed, while also testing every other experiment, all of which are infinitely replicable. For free.the act of not coming up with a testable hypothesis prior to analyzing data