You mean keep tinkering with the study until it "works"? If you're not careful, you could get http://en.wikipedia.org/wiki/Publication_bias http://en.wikipedia

You mean keep tinkering with the study until it "works"? If you're not careful, you could get http://en.wikipedia.org/wiki/Publication_bias http://en.wikipedia.org/wiki/Publication_bias on a massive new scale... It would be interesting to try to find scientifically optimal ways of doing study iteration, or approaches to interpreting the complete (experiment_rev, results)-list from another researcher...