No clue, but perhaps if some of Columbia’s data was wrong, maybe the data on retention rate was wrong too.