Hello Hirsch

Humbug, Hirsch’s h-index! The humbling yet heartwarming hooligan, happily hopping homeward.

Many people have issues with the widespread use of the h-index in determining one’s scientific worth. But the h-index is not inherently bad; it simply needs to be considered in the context of one’s field and seniority. Even in a given field, the h-index is higher for the hot, fast-moving topics than for the slower ones. Also, it tends to be high in larger fields and, more generally,  in fields in which the number of papers published per unit time is high. For instance, brilliant mathematicians will have much, much smaller h-indices than brilliant biomedical engineers or chemists. While some people like to compare the h-indices too broadly, between fields with vastly different dynamics, as it makes them feel superior or something, that’s just silly; I think (hope?) that most reasonable people understand this fallacy in the use of the h-index.

However, I want to discuss the use of the h-index for very junior folks. The h-index can certainly emphasize that an individual is unusually creative or high-impact, but let’s face it: unless we are dealing with math or theoretical physics or another endeavor where the intellectual labor of a single individual  spawns a paper, it’s the teams, and sometimes very large ones, that are behind each paper in the physical and biological sciences. If you are in a productive, successful group, where people are not particularly stingy about co-authorship, you can achieve a pretty high h-index largely by association.

A couple of years ago, we were selecting from among several candidates for the department best-dissertation awards. The award went, against my protests, to a young person who had the highest h-index. Mind you, the h-index was high because this person did research in a great group as an undergrad, and had a number of papers where they were author number 7 or 8 among about 15; then they were a member of a team in grad school, where again there were many papers with the person author number 4 or 5. Now compare this with the work of someone who had 5 or 6 first-author publication in the last two years of their PhD. These papers were fewer in number and likely not highly cited at the point of graduation, owing to the short time (most just got out when we considered the award nominees). In contrast, the well-plugged-in first person had a fairly high h-index, as they were part of strong and prolific groups for a long time, making a minor contribution to each paper, but the papers were more numerous and out for longer and had the time to get cited (in case you are wondering, that person only had one first-author paper).

Anyway, h-index is not evil, it’s a just number.  It’s the people who can be misinformed, lazy, or simply self-serving and manipulative in how they utilize it.  Then again, people would abuse any other available metric as well; at least this one is transparent and easy to calculate.

(Btw, everyone should make a Google Scholar profile. It’s very pretty. And people google your citation count anyway, you might as well have some control over what they see.)

5 comments

  1. Yeah, google scholar is the bomb! I just got mine set up.

    The only “problem” is an early paper of mine (which honestly is about something I don’t care much about anymore) has about a billion times the citations of the work I do care about so it sticks out like a sore thumb on my google scholar… It was a new methodology in a superhot (at the time) field…

  2. Yup Google scholar is awesome and much more comprehensive than other such services ! Also I can totally control (almost) all things on it so it works out great for me. Some people prefer ResearchGate which has its own impact score which is suppose to eliminate all h-index fallacies. I dont know how effective it is though.

  3. Juicing up h-index by being a middle author on tons of papers is a big issue for me.

    Where I am at (not US), the culture is that everyone just counts the number of papers and the impact/citations without bothering to look at actual individual contribution. This is very very evident during grant appraisals and promotions.

    As a junior prof, I am really struggling to identify the right balance between contribution (that is, my last author papers) vs collaboration (middling author positions). My personal goal is that at least 60-65% of total publications will be from my own group. I have been told that this is too high and ambitious, may be seen as an indication I dont play well together.

    I absolutely hate this from the core of my guts. I wish the culture here was like in US where only the last author papers count. But your post suggest that you may be losing it too. Sigh.

    I need to stop here before I get too angry and start whining about people writing funcken review papers just to pump up their citations.

  4. As someone that uses numbers all the time to make all kinds of decisions, I hate it when people just look at numbers to make decisions! What other option is there? As people are more selfish and/or have less and less free time, looking at a number is a quick-and-easy way to widdle people down. For yearly reviews we started adding five numbers with slashes to describe the employee’s whole rating. This told a better story of the person’s year. But, of course, some HR dude thought it would be cool if we just average all the numbers. So now I just get one number…

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s