I have been running a little experiment on twitter. Oh well, it doesn’t really
deserve the term “experiment” – at least in an academic vocabulary – because
there certainly are no treatment effects or control groups. It does deserve the
term “little” though, because there are only four observations.