Publishing less: our current thinking about comparative statistics