There is an excellent comment on research misconduct at the brand new Pubmed Commons site by Dorothy Bishop:
Instead of valuing papers in top journals, we should be valuing research replicability. This would entail a massive change in our culture, but a start has already been made in my discipline of psychology (see http://www.nature.com).
… The top journals prioritize exciting results over methodological rigour. There is therefore a strong temptation to do post hoc analyses of data until an exciting result emerges. I agree with Farthing that pre-registration of research projects is a good way of dealing with this. I’m pleased to say that here too, psychology is leading the way in extending research registration beyond the domain of clinical trials (http://blogs.lse.ac.uk/).
That’s the way to go.