Talk:Days since first edit

This is fascinating! One observation: the introduction to the page says "samples were chosen by taking one edit per 10 minute period." This could be quite different than what would result if editors were randomly sampled; specifically, the current approach leads to overrepresentation of people who edit frequently. I would also love to see how the results would look if editors were randomly sampled instead of edits. Tobacman 14:06, 6 June 2006 (UTC)[reply]

How do you want to sample editors instead of edits? If you sample from all editors then you should get distribution of wikipedians by time they started to work - this is already available at http://stats.wikimedia.org/EN/ -- Nichtich 13:26, 7 June 2006 (UTC)[reply]