All That Matters

As this week there is again a lot of talk about journal impact factors with the release of this year’s data later today, I like to take this timely opportunity to look at citation metrics more broadly, in terms of fundamental flaws in weighing data, and important data missing in the underlying data sets, which in my view miss important data when it comes to practical, technological impact of a study.

I recently had the opportunity of attending a talk by Paul Wouters from Leiden University, a professor of scientometrics. He pointed out one of the fundamental flaws in citation metrics that goes right to the heart of such data collection, before one should even discuss more superficial metrics such as h-index or the impact factor. Like any other piece of data, the context of a citation matters, he said. Factors that play a role are the type of paper where a reference is cited, and in what way. Was it criticism? Controversial papers for a while at least can gather a lot of citations even though eventually their impact on scientific process can be nil. There are also human aspects. Relevant points here are who cited a paper, was it a self-citation, or were there other motivations for citations? After all, citation cartels are not unheard of.

There is a lot of literature on various aspects of citation analysis, and more details on this can be found in Wouters’ doctoral thesis on citation culture, or in the 2008 paper by Jeppe Nicolaisen on citation analysis.

More broadly speaking, I am not sure whether it will be possible to properly analyse and process context when it comes to citation analysis. There are too many ways to game such systems. However, a more complex analysis might well be possible, taking the example of he ranking of web sites in search engines. There, context is everything. A website that is linked from many other sites is not necessarily an important one. Instead, a link to a web site from an important web outlet such as a popular news web site weighs much more than links from unknown web sites. Indeed, many links from news web sites or social networks might also be an indicator of immediacy, further propelling a site up the search engine rankings. […]

I’m glad I’m travelling this week, as yesterday the impact factors have been announced. I actually forgot about the pending announcement yesterday, when some of my hosts here in Singapore reminded me about it at dinner(!). And sure enough, last night twitter was all abuzz with impact factors.

The reason I am happy I’m away from all the buzz is that the impact factors tend to get quite overemphasized. I mean, really, what does it mean, ‘impact’ factor. Per definition, the 2010 impact factor counts the citations in 2010 to papers published in 2008 and 2009, divided by the number of those papers. For more details, there is an explainer by Thomson Reuters, who publish the Impact Factors.

So what kind of impact does it measure? That of papers published there? Well, citations to papers vary a lot. Take my journal, Nature Materials. The most cited paper from 2008 for example is a review on Biosensing with plasmonic nanosensors that has 473 citations so far – coincidentally, that’s a review that I commissioned, and sure I’m glad I did. Going further down the ranks of the most cited papers from 2008, the paper at 10th place got 175 citations so far, that at 20th place 104. In other words – the impact factor has not much to do with the quality of a single paper, the distribution of citations varies a lot.

Therefore, the impact factor certainly doesn’t measure the ‘impact’ of papers, and by implication nor does it measure that of researchers. It provides an average number for a journal. But that even that doesn’t look like an absolute measure to me either. The impact factor of Nature Materials is 29.897 (yes, it’s calculated with such silly accuracy). In comparison, that of our sister journal Nature Physics is 18.423.  So does that mean that Nature Materials is 62% better than Nature Physics? You better start reading the journals, as that’s certainly not the case. Of course, a journal like Nature Materials might be perceived to publish papers on average better than some journals with an impact factor of say below 10. But as a researcher you would have already known that from reading the papers published in a journal, wouldn’t you?

So what use is the impact factor number? Well, being cynical one could say it is a quick measure for those that don’t read the journals but still want to know how good they are on average. The danger is of course that this is then used as a kind of metric to assess the quality of research or to decide on the career of researchers. As it’s clear from the examples above, it certainly should not be used for that purpose.There are better ways to judge the merits of published research, such as article-based metrics and not journal-based ones. Not even my salary as an editor depends in any way on the impact factor of the journal I edit, so certainly it shouldn’t impact those of researchers.

And that’s also because the impact factor is a woefully short-term metric. On Monday at the conference here in Singapore I listened to an interesting talk by Jonathan Adams from Thomson Reuters, and he showed a citation statistics whereby for most disciplines in the physical sciences the number of citations to papers steadily increases over the years until it reaches a maximum at around 12 years (give or take a few years). So even on average for all publications considered, measuring citations for only the last two recent years can mask the true impact of a paper.

Where such short-term metrics can be useful, however, is as evidence for considerable editorial efforts by a journal. For example in case of the remarkable 30% increase in impact factor for Nature Materials’ competitor, Advanced Materials, whose impact factor now stands at 10.857. My congratulations to them on their hard work! But all in all, we shouldn’t overplay the relevance of impact factors.