Sisk et al.’s Scholarly Impact Study and Mean Minus Median Scores

Following up on a couple of posts last week on Gregory Sisk et al.'s  "Scholarly Impact of Law School Faculties: Extending the Leiter Rankings to the Top 70", I want to look some more at the differences between schools' mean and medians.  My last post presented a table that ranked schools based on mean and median and pointed out that some schools had rather dramatic differences between their mean ranks and their median ranks.

Now I want to look at absolute differences between schools' mean and medians. This table (which uses the latest version of the Sisk data) ranks schools based on the difference between mean and median (not the difference between the ranks, but the difference between each school's actual mean and median).  Incidentally, all schools had a higher mean than median.  Some schools — especially those near the top of the Sisk/Leiter ranking — have substantially higher means than medians, which indicates that those schools have some scholars who are gaining extraordinary numbers of citations, even among their already highly-cited faculty.  One other thing that's pretty interesting is the high correlation between the difference in mean and median and the weighted score (.85), which indicates that the schools that do well on the weighted score (higher is better) also tend to have a larger difference between mean and median than schools with a lower weighted score.

One other point: Sisk's top 70 schools now includes 71 schools.  I'm hoping to get some of my previous tables re-run shortly.

The illustration is the Institute for Scientific Information's building on Market Street in Philadelphia.