Conversation or vocabulary? It seems to me they conflate the two simply because conversation time is correlated with higher vocabulary. Do the studies control for this?