Reminds me of a friend of mine who once asked me to help her with a test for statistical significance of some data of her PHD thesis* in micro biology ...

When I tried to explain the procedure to her she quickly gave up and told me the problem was already solved.

Unfortunately I insisted to know how it was "solved". Turned out that her department used an excel formula for these tests which changed some colour from red to green.

They achieved it by permuting the data from cell to cell until something turned green and than published it right away.

*) actually it wasn't her data, part of her contract was to help preparing publications for colleagues in London...