Unreliable research: Trouble at the lab. Science relies on the replication of experiments. The concern is that a startlingly small number of published papers hold up to attempts to reproduce the results. The Economist looks at what that means for the field and for academic research more broadly.