There’s far more scientific fraud than anyone wants to admit

The solution might be to shift from peer review to peer replication.