Clarivate also makes the Eigenfactor score, the Article Influence Score, and they just came out in this year with the Journal Citation Indicator. They also have a Cited half-life and an Immediacy Index. Their competitor Elsevier has some metrics including the Cite Score, SCimago journal rank or SJR, the SNIP (Source Normalized Impact by Paper), and they have the h-index. (See previous post on the h-index.) The h-index metric that is used for individual researchers, but we can also make an h-index for journals in a very similar way. Google Scholar has a couple of metrics that are called the h5 index and the h5 median. For definitions, I found this really helpful web page on the Journal of Experimental Biology where they list most of these metrics and have their definitions.

Another interesting metric is the Altmetric, which measures the media impact of scientific publications. It includes mentions in news outlets, in blogs, in policy source tweets, Facebook pages, Wikipedia pages, Google Plus, Reddit, and videos. It’s quite a bit different than looking only at citation rate.

Let’s take a look at how some of these metrics compare to each other and to the Journal Impact Factor. This first image contains the Journal Impact Factor for 15 well known journals that are ranked from very high impact factor to moderate impact factor. We also see the Eigenfactor score, the h5 index, the h5 median, the Cite Score, the SNIP, and the SJR. Now, they’re still sorted in order from highest impact factor to lowest, color coded these with a heat map, so you don’t have to read all the numbers if you want to just look at the colors. It is clear that there is not an exact relationship. The highest ranked journal impact factor does not necessarily have the highest h-index or h5 median, for example. And the Eigenfactor score is kind of all over the place.

Next I converted these metrics to a ranking within these 15 journals and I re-sorted the order of the journals by the mean of the ranking, which you see in the far right hand column. Now instead of the highest number being the best, now the lowest number is the best, because rank number one is the best. For many of the journals, the order has changed compared to the Journal Impact Factor alone. For example, Nature was sixth on the journal impact factor, and now it’s in second position after averaging across all of these metrics.

The color pattern is starting to become a bit more similar between the columns. The order is not exactly the same, but the red ones tend to be at the top and the blue ones or the lower rank tend to be towards the bottom. The one metric that is really not correlating all that well with the other ones is the Eigenfactor score. So I decided to take that column out and re-rank everything.

Now it is a much more regular pattern, but again these metrics don’t match up in a one-to-one basis. So, it is important to take into account more than one of these metrics when you are evaluating a journal. But we can clearly see that there are certain journals that are in the highest tier, journals that are in the mid-tier, and journals that are in a lower tier. Overall, I think that is a better way to rank journals than a fine scale. By looking at multiple metrics, it becomes harder to say that one journal is better than another because it has a Journal Impact Factor that is one or two numbers higher.

The take-home message is: don’t rely entirely on journal impact factor. In fact, don’t rely on any single metric. Overall, these metrics do correlate somewhat to each other, but not exactly, and  it’s important to take into account multiple pieces of information when you’re evaluating which journal you want to submit your paper to.

Do you have any additional thoughts on journal metrics? Leave a comment below!