How much of science is reproducible? It depends how you approach the question, but some papers have found that more than half of studies in psychology fail to replicate. Other social sciences aren’t much better.