Impact factor: academic quality or academic snobbery?

Twitter screenshot - retracting article due to decreased impact factor

I stumbled across a tweet over the weekend that prompted me to write this blog article. As shown in the image, the author has withdrawn an article they wrote because the journal’s impact factor has fallen, which got me thinking: Why are academics so obsessed with impact factors?

Firstly, what is impact factor? Impact factor is the yearly average number of citations, or how many times an article is referenced in other articles in the past two years. It is frequently used as an indication of the quality of a journal within its field. A journal with a high impact factor is often seen as carrying more prestige than ones with a lower impact factor.

If we look at Nature’s impact factor 2017 (taken from Wikipedia), we see that the impact factor is 42, which means that, on average, papers published in Nature in 2015 and 2016 received roughly 42 citations each in 2017.

{\displaystyle {\text{IF}}_{2017}={{\text{Citations}}_{2017} \over {\text{Publications}}_{2016}+{\text{Publications}}_{2015}}={74090 \over 880+902}=41.577}

Impact factor is contentious, and there are several arguments against its use as a metric. Firstly there is an issue of comparison across disciplines; the total number of citations in subjects such as physics or maths is often lower than in the biosciences. Therefore the impact factor between disciplines is subjective. Impact factors can be heavily influenced by a few very highly cited papers published in the journal. In the example of Nature above, very few of the articles published between 2015 and 2016 will have been cited 42 times, but there are likely to be a handful of papers that have been cited very highly, which highlights the flaw of applying one metric to a highly skewed distribution of papers.

Furthermore, impact factors are often used as a proxy for paper quality, but a seminal paper in a field such as microbiology might not be within the remit of journals with high impact factors. Does that somehow devalue the quality of this paper…well, no, it’s still a seminal paper, and I think we desperately need to move away from impact factor equals quality.

Journal editors are also charged with maintaining or enhancing their journal’s impact factors and, as such, can be very selective about what articles they accept to maintain these numbers. A slightly more underhand approach is to ask potential authors to cite articles from the journal that wish to publish in, thereby artificially inflating impact factors.

Historically impact factor has been used to judge the quality of an academic’s output. Thankfully many universities have signed up to the San Francisco Declaration on Research Assessment (DORA), which aims to end this practice and instead judge academics on their individual outputs.

So back to the original question, quality or snobbery? Well, I don’t see anything wrong with academics wanting to publish in a high impact journal, but that impact factor can be used to judge the overall quality of a journal within a specific discipline and should definitely not be used cross-disciplinary or to evaluate the quality of individual papers.

At the end of the day, if your work is deemed worthy of publication anywhere, it is impactful and adds to your subject’s knowledge base. To then withdraw a paper later because the “impact factor” of that journal changes deprives others of that knowledge and is nothing more than academic snobbery.