Williams et.al article at The Conversation

Does publication bias make antidepressants seem more effective at treating anxiety than they really are?