The largest-ever database of retracted articles suggests the burgeoning numbers reflect better oversight, not a crisis in science.