The Hirsch Index (or in a sort way, the h-Index) is a way to measure scientific popularity by one number. A scientific with an h-Index of n, will have n papers with at least n citations each one. Note that I am saying that it can measure "scientific popularity", neither "scientific excelence" nor "scientific quality" (although there is often a correlation between those three facts).
h-index is sometimes a way to self-glorification, other times it hides a collaborative mafia ("I will cite you if you cite me"), but I like it (and so scientific community do). It is only a metric, but, as in other metrics, is a quantitative way to measure the importance of a scientific in its community.
As you can see, h-index grows in an exponential manner: when you get your first citation, you get your h=1. To get h=2 you need, either to get (al least) one more citation on that paper, and (at least) two more in a different one, or get (at least) two citations in different papers. That means that, stepping from an h-index of n-2 to n-1 is quite easy than doing the same from n-1 to n (because in every step you need more and more citations).
A tool which is helpful to compute this index is "Publish or Perish", available here. Using google scholar and other similar services, this tool can easily compute the h-index of a researcher, in a semi-automatic way (you often have to discard manually several publications that are not coming from that author, and of course, self-citations).
By the way, at the present day (2/12/2008) my h-index is 1. It is not so bad for a PhD student, but I hope it would be improved next year...
h-index is sometimes a way to self-glorification, other times it hides a collaborative mafia ("I will cite you if you cite me"), but I like it (and so scientific community do). It is only a metric, but, as in other metrics, is a quantitative way to measure the importance of a scientific in its community.
As you can see, h-index grows in an exponential manner: when you get your first citation, you get your h=1. To get h=2 you need, either to get (al least) one more citation on that paper, and (at least) two more in a different one, or get (at least) two citations in different papers. That means that, stepping from an h-index of n-2 to n-1 is quite easy than doing the same from n-1 to n (because in every step you need more and more citations).
A tool which is helpful to compute this index is "Publish or Perish", available here. Using google scholar and other similar services, this tool can easily compute the h-index of a researcher, in a semi-automatic way (you often have to discard manually several publications that are not coming from that author, and of course, self-citations).
By the way, at the present day (2/12/2008) my h-index is 1. It is not so bad for a PhD student, but I hope it would be improved next year...
1 comentario:
1 is not so bad. In any case, H is difficult to pervert, at least beyond a certain level. By roundtable-citing cliques you might obtain maybe H=5, or 6, or even 10. But to go beyond that, first, you need to have more than 10 papers published, and then, you need to have them cited.
I would say, anyways, that there is certainly a correlation between H index and, if not quality, kwalitee. You get a high h index if your paper filled a niche, or if it came in a particular moment when it was needed, or if you have many collaborations around the world and thus many different lines of research. Of course, your H increases if the surname of the first author of a papers starts with a A in a field where alphabetical ordering of citations is the rule. In any case, it's not easy to get an H of 20 or 30.
Publicar un comentario