As I mentioned last week, I decided that there wasn’t enough interesting about controlled vocabulary to fill up a whole month of posts so I’ve decided to switch topics for this week and next to journal metrics.  The first metric I’m using in some research I’m conducting with a professor, and it’s called the H-Index.

H-Index was originally introduced by J.E. Hirsch as one way of measuring an author’s impact on his field.  The way it works is you order the author’s papers from the most to least cited paper along the x-axis.  Then you graph on the y-axis the number of citations for each paper.  You then draw the line x=y.  The intersection of that line with the line graphing the papers and citations is the H-Index.  So the H-Index gives you the number of papers (n) that have been cited at least n times.  Both Web of Knowledge and Scopus will calculate H-Index for you, and there are various plugins you can use with Google Scholar.

In the example above, this author has written more than 60 papers, some of them have never been cited (not unexpected), but one of them has been cited almost 60 times.  However, his H-Index is 16 because he has 16 papers that have been cited at least 16 times (including the one that has been cited almost 60 times).

Most authors who are reasonably productive will have a graph similar to the one above.  But it is important to remember that H-Index is very context specific.  Presumably more seasoned scholars will have higher H-Index than younger scholars (they have more papers and have had more time for their papers to be cited), and the H-Index “standard” will vary with different disciplines.  Nevertheless, H-Index is growing more popular than just the simple number of publications or the number of total or average citations because it tends not to favor one-hit wonders or people who publish a lot of drivel.

In light of the information above, our research is going to investigate what a reasonable H-Index should be for an assistant professor in his field going up for tenure.  We are also interested to see whether one can “rank” departments based on a cumulative H-Index for all the professors in that department.  This field has both research and clinical components, and we expect that the H-Index for departments that specialize in one area or the other might be quite different.

So, what’s your H-Index?