On the meaninglessness of impact factor

Original post at Twitter here

Given chatter about impact factor lately, I got thinking y’day about how much the average scientist should care about it at all…

Scraped my Google Scholar to see the correlation between my paper citations and the (current) JIF of the journal. There isn’t one (R² < 0.03)

Average citations per year vs journal impact factor, for me

I’m not surprised and it’s probably obvious, but I have heard many times impact factor come up in the context of deciding where to submit. It’s something I have always tried to ignore, though it’s probably part of why I don’t make a very good academic…

Second, have seen a lot on Twitter from journals and editors happily announcing impact factor increases this year. Happy for success of course, but I wonder how much is down to an increase in publishing activity in general.

I went to WoK and looked at citations per year for whatever search term I could think of. Really hard to find anything where citations are not rising exponentially. Basically the only things that weren’t were specific searches for obsolete technology e.g. “floppy disk”.

Citations per year for some random search terms

I think @CBommier’s tweet rings very true. Basically everyone when pushed would agree JIF is no measure of quality of an individual researcher or paper. But I am sure very many see JIF as a measure of prestige which factors into a lot of decision making.

Most active scientists, I am sure, are PhDs and postdocs, the vast majority of which will not become professors and instead go to non-academic careers, where basically nobody cares about JIF.

Yet the pull of publishing in “high impact” journals still means many feel the need to aim high, spend months or years rolling the dice with different editors/reviewers, and all the stress that comes with that, when for most PhDs/postdocs it probably doesn’t matter at all.

Since I don’t work in academia now, I think about JIF even less than before. But nonetheless I hope science/academia finds an alternative to all this metrification, since I hardly see any benefit except to those lucky enough to be able to use it to their advantage.

Meanwhile my position is: share preprints and give people the chance to read - I don’t feel I need peer review to defend the quality or relevance of my work - and submit to journals which are relevant and hopefully have a good community supporting it. Down with metrics…

comments powered by Disqus