- Aggregate rankings are based on a harmonic mean of rankings for each criterion. Those individual rankings differ between worldwide and regional groups. As an example, an author who is much more cited than his regional peers is ranked far ahead in the worldwide rankings, but no such gap would appear in the regional rankings.
- For authors with multiple affiliations in different regions, only part of their score will count towards their regional ranking. See above how those weights are computed, or see this blog post.
In any case, the ranking as it would result from an exerpt from the world ranking is listed in square brackets next to each regional score. Note that their maybe rankings reversals between the regional and worldwide rankings. To see this, take an example of two economists from the same region and seven ranking criteria:
|Economist A: || 1, || 2, || 3, || 4, || 5, || 6, || 7, |
|Economist B: || 7, || 7, || 7, || 3, || 4, || 5, || 6, |
As the top and bottom ranks are dropped, the harmonic mean from (2, 3, 4, 5, 6) is 3.5 for Economist A and 5.5 (from (7, 7, 4, 5, 6) for Economist B. In the regional ranking, assuming these are the two top economists for each criterion, A has (1, 1, 2, 2, 2) for a mean of 1.43 and B has (1, 1, 1, 2, 2) for a mean of 1.25. Thus A is ahead in the world ranking, and B is ahead in the regional ranking. This happens because A is much ahead of B for some criteria and little behind for some others in the world ranking.