Beware representing a distribution with a single number. Individuals disappear entirely and it's easy to make incorrect generalizations. Your conclusions don't

Beware representing a distribution with a single number. Individuals disappear entirely and it's easy to make incorrect generalizations. Your conclusions don't necessarily follow from your statistics, and I think this is something important enough that rigor is necessary.