[glitchc]

The OP's point is, I'm afraid a valid one. Attempting to find a statisticallly significant to report on in time to meet a conference deadline means experiments are fudged, results are cherry picked and discrepancies are swept under the rug. Conference and funding deadlines coupled with the need to generate a publication record are almost entirely responsible for the dubious nature of the average scientific article. It's a classic case of Goodhart's Law.