Thursday, March 10, 2011

Correction


Apologies for repeating a bad meme -- according to which Prof. Venki Ramakrishnan, one of the 2009 Chemistry Nobel winners, fares poorly in citation metrics.

I first found that meme in a comment on my blog way back in 2009. I had also heard it mentioned in conversations on quite a few occasions later. When I found a version of it again in this Current Science letter by my colleagues Prof. Ramasesha and Prof. Sen (I must hasten to add that their wording is a lot more careful, and they don't mention any numbers), I thought it was worth highlighting.

Then Giridhar left this comment; and Sunil followed up with another comment yesterday.

A search in Web of Knowledge (whose database is arguably the most comprehensive) reveals the essential correctness of Giridhar's comment [I couldn't verify the exact numbers he quoted because I'm not sure about my own search skills at WoK].

My search for a (possible) source of the bad meme led me to this site which calls itself Microsoft Academic Search; it occurs at No.2 in this search.

Lessons: Trust but verify -- as far as possible, and however authoritative the source might appear to be. And issue a correction when a mistake is discovered.

Apologies.

And, thanks to Giridhar for pointing out the mistake.

2 Comments:

  1. Anonymous said...

    Here is the original PNAS article proposing and discussing the h-index:

    http://www.pnas.org/content/102/46/16569.full

    It also contains a distribution of h-indices among Nobel laureates in physics.

    However, what would be more interesting is the h-index distribution across ALL scientists.

    And that is here, http://www.brics.dk/~mis/hnumber.html

    This should settle the question of whether Dr. Ramakrishnan and others are "good enough" by h-index standards. For instance, there are about a 100 scientists in computer science with h-index greater than 40.

    http://www.cs.ucla.edu/~palsberg/h-number.html

  2. Giri@iisc said...

    Well, I should said how the search was conducted on web of science. Here it is, so that it can be reproduced by others.

    Author=(ramakrishnan, v)
    Refined by: Institutions=( BROOKHAVEN NATL LAB OR MRC OR UNIV UTAH ).

    Papers: 99, h-index:47 and citations:8888

    A h-index of 47 or citations of 8888 are not low by any standards. In fact, it directly qualifies him to be in the top 1% of all scientists in chemistry. There are other scientists who have also published very few papers but the number of citations are very high (Sanger published around 60 papers with >90,000 citations and got two Nobel prizes).

    However, h-index or number of citations is not the criteria for fellowships or awards. It is just a parameter that can be used judiciously to look at serious scientists. It is unlikely a scientist whose number of citations is less than 100 is going to win the Nobel. It also does not mean that all scientists with more than 8000 citations are going to win the Nobel.


    Of course, in India, the criteria for judging an applicant is based on the very objective parameter called the quality of mind.

    Thanks

    Giridhar