Hi Martyn
It's quite prevalent. I saw it most recently in behavioural psychology.
...
I've also seen it many times in other areas of science. I can't recall the details: I'm not an academic researcher!
PS: Needless to say, I'm not saying that it's true for ALL academic research. It's a completely understandable phenomenon. If I do a study and conclude that roast beef doesn't cause cancer, no-one will notice, but if I can find some way to adjust the statistics or the statistical significance test I'm using, to conclude that it DOES cause cancer, then I'm onto a winner in terms of publicity.