Small samples?

The ABC’s top 100 book list has four ties (and they don’t skip the next number, so there are 104 books in the top 100). The bigger the sample, the less likely there will be ties. The ABC says that "tens of thousands" of people voted, but my guess is that once you omit the top 10 books, the numbers voting for the rest would be pretty small. Anyone care to have a crack at the maths?

This entry was posted in Uncategorized. Bookmark the permalink.

2 Responses to Small samples?

  1. Bryan Palmer says:

    The sample size might be much larger than you think.

    For something comparable, take a look at this site listing the most cited blog articles for 2004, http://www.blogpulse.com/2004_review/2004review_blogposts.html.

    The most cited article had 3089 citations. The 50th most cited article had 189 citations. There are more than 21,000 instances of citation in the top 50. There are three ties in the list 26 and 27, 31 and 32, 49 and 50.

  2. Andrew Leigh says:

    Bryan, useful piece of evidence, thanks. Love to see the maths, though….

Comments are closed.