Using twins as the anecdotal sample for whether children have best friends is another point against the thoughtfulness of the author. As a twin myself, I would argue that twins have a built-in best friend and thus rarely have another "best friend" like a singleton would, so they are a particularly poor choice to illustrate anecdotal evidence that kids don't have single best friends.