Tuesday, July 31, 2018

Should I be proud of my h index?

My daughter has a hoodie that says: “Don’t let the number of likes define your art”. The basic point, of course, is that quantitative measures of popularity are not how a person’s work should be valued by others – nor especially by that person. Rather, it should be the quality of the work that person produces or – more fundamentally – the quality of the character of the person. Thus, the large number of followers that the Kardashians, or some Instagram star or Youtuber, has should not be used to evaluate how creative, intelligent, or innovative the person is – or how good of a person they are.

In academia, the same concerns arise because the formal or informal valuation of a person’s research, and therefore that person as a scientist or even as a person, is sometimes tied to quantitative metrics not so different from the “number of likes” or the “number of followers.” Of particular significance are the number of papers we publish, the number of citations they receive, and – together – the h index they generate. The immediate motivation in using these sorts of metrics for evaluation is that they provide an objective way of comparing different academics, at least within a discipline, to thereby determine their relative worthiness for grants or promotions or endowed chairs or raises. The broader hope, of course, is that the same metrics reflect not only the quantity of research but also its quality.

Recently, these social media and academic quantifiers of success and influence have converged in the consideration of the number of followers that a given academic has on twitter or Instagram or youtube or whatever. And papers now receive an "Altmetric" score reflecting its social media attention. This convergence has led some to propose a “Kardashian index” (or k-index) measuring the extent to which the social media profile of a scientist (their followers on twitter) is an outlier with respect to their academic profile (their total number of citations).

This frequent use of quantitative metrics of influence for evaluation has received many criticisms – but the most important is that volume (papers) and popularity (citations) do not reflect the quality of a person’s work. Instead, they can mostly reflect an individual’s effectiveness in self-promotion, most obviously in the form of self-citation. (See my blog post on the “Narcissist index” or N index.) Or they could reflect the tendency of that person to write review papers, which tend to be more highly cited. (See my blog post on “What if all my papers were reviews.”) Or they could largely reflect the discipline in which a scientist works (See Dan Bolnick's companion post on "Do certain subdisciplines lead to a higher H index".) The main argument is that quality is instead a more subjective evaluation achievable only by reading the person’s work and qualitatively placing it in the context of other work in the field.

These criticisms of quantitative metrics of influence in academia are so frequent and strong that it makes one wonder “should I be proud of my h index?” For instance, I can see that – in the not-too-distant future – I will have 200 papers, 20 000 citations, an h index of 100, 5000 followers on twitter, 1 000 000 views of this blog, and 250,000 views on Youtube. Should I be proud of these numbers? Should I report them in grant proposals? Should I be embarrassed to enumerate them in this blog post? The truth is, I don’t know. I have mixed feelings about it. I am proud to see that my hard work and talent for self-promotion has translated into a substantial visibility in my field and also effective communication of science to a broader audience. At the same time, I recognize that these numbers are not necessarily an indicator of the quality of my work, as they would undoubtedly be lower for someone who wrote fewer reviews, did less self-citation, joined fewer perspective papers, took fewer students and put their name of fewer of their student’s papers, and so on.

So perhaps, as I tell my daughter, her shirt might be modified to say: “Don’t let the number of likes define your art … unless maybe you have a lot of likes.” (Of course, “define” is not the optimal word either.) Maybe, in this sense, it is defensible to be proud of my quantitative numbers simply for their own sake, while recognizing that they reflect little more than the fact that they are high numbers. Stated another way, having a lot of citations and a high h index should not be a scientific goal but rather a communication goal. From this perspective, it is reasonable to appreciate and reward someone with a high h index or many twitter followers, without any sort of judgement being placed on someone with lower numbers. Many exceptional scientists who are influencing their field do not have a high h-index – but that does not make they any less important or any less of a scientist.

I encourage everyone to consider social media presence and citation rates and so on as indicators of nothing more than social media presence and citation rates. If you value such things for themselves, great – be proud of them – but realize that those numbers don’t necessarily reflect anything else. If your goal is to have high numbers, then – yes – write lots of papers, take lots of students, and write lots of review papers. If your goal is to be a good scientist; by all means do those same things if they lead you to do good work – but not if they detract from the quality of your work. And, most importantly, don’t let the pursuit of papers and citations stop you from being a good colleague, collaborator, and person. And, most of all, never let any of this come in the way of family and the things you value as a high quality of life.

Don’t let your h index define your science … but you can be proud of it nonetheless.


  1. Another great post. I do think H index is one of the better ones... Overall influence is a product of quantity and quality.

  2. This comment has been removed by the author.


A 25-year quest for the Holy Grail of evolutionary biology

When I started my postdoc in 1998, I think it is safe to say that the Holy Grail (or maybe Rosetta Stone) for many evolutionary biologists w...