New Improbable Research Blog Entry: "Galam’s Work on Galam Models (Reviewed by Galam)"

Here is my latest post for the Improbable Research blog. It is called "Galam’s Work on Galam Models (Reviewed by Galam)".

The Most Prolific Scholar Ever

With more than 2.4 million citations and an h-index of 333, this prolific scholar puts even Paul Erdős to shame. :)

(Tip of the cap to Ramis Movassagh.)

Update: Peter Mucha points out that it's worth following the link to this prolific author's homepage. I didn't do that before; Peter is absolutely correct.

Ranking Scientists According to Sarcasm

"Instead of using citations, I think we should rank scientists from most sarcastic to least."

Naturally, I was accused of using my own CV after I proposed that criterion for measuring quality.

Maybe I should call this metric the "p-index"? We could call it the "s-index", if you prefer, but then I wouldn't be naming it after myself.

Self-Citations of Men Versus Women

According to a new study, male scientists self-cite more than female scientists. One of the authors of the study is Carl Bergstrom, so naturally I am wondering if the paper includes a citation of a Rosvall–Bergstrom paper on community detection.

(Tip of the cap to Oxford Mathematics Good Practice.)

Headline: Cambridge University Press Breaks Google Scholar

Congratulations, Cambridge University Press! You have broken Google Scholar. I noticed that a paper of mine (Wymbs et al., Neuron, 2012) suddenly went from 40 citations to 70 in one Google-Scholar update. No other paper jumped so crazily this time, so I ruled out an algorithm update. It turns out that there is a recent 'target article' in Behavioral and Brain Sciences that includes much invited 'open peer commentary'.

One comment cited us, but each comment/article has a separate DOI even though they all link to the same .pdf file. 30 instant citations.

Now I wonder how much BBS inflates their impact factor by doing this... (Important note: this effect can easily be an unintended byproduct. I am just pointing out that it does indeed appear to be such a byproduct of how the website has organized the article and its associated comments. I am making no comments whatsoever about anybody's intent, because I have no idea about that.)

I am amused. (Also horrified.)

Where Citations Come From

Here is a new xkcd strip about where citations come from. They're kind of like babies but slightly less messy.

Citation Errors as Genetic Mutations

As this opinion piece discusses, citation errors are rather analogous to genetic mutations.

Unfortunately, the optimism expressed at the end of the article that referencing software will help matters is misplaced---my experience is that such things induce new errors rather severely.

(Tip of the cap to Julius Su.)

Impact Factors Suck

I've said it before and I'll say it again: Impact factors suck. They can be gamed, and treacherously so at that, and they have little to no bearing on the scientific quality of a journal or the articles therein. Read this article for some examples, and indubitably there are numerous smaller (and thus harder to catch) bad practices all over the place. We need to be more discriminating in what journals we consider good, and I feel like screaming every time I hear somebody using a journal's impact factor as a reason to submit there. (And in some fields, such as pure mathematics, the citations often don't tell the real story anyway!) And I feel like screaming even louder whenever I hear about impact factors of the journals in which in author is published being used for hiring, promotion, tenure, or anything else that can be very meaningful to people.

When I publish a paper, I try to choose journals whose papers I actually read or that I feel will reach the audience I want to reach---independent of numbers, etc.

Let's spread the word!

(Tip of the cap to Predrag Cvitanovic.)

Tales from the arXiv: Impact factor edition

Here is an abstract, just posted on the arXiv, that attempts to analyze the inflation in journal impact factors over the years:

arXiv:0804.3116
Date: Sat, 19 Apr 2008 00:17:22 GMT (995kb,D)

Title: Differences in Impact Factor Across Fields and Over Time
Authors: Benjamin M. Althouse, Jevin D. West, Theodore Bergstrom, Carl T.
Bergstrom
Categories: physics.soc-ph
Comments: 9 pages, 3 figures
License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
\\
The bibliometric measure impact factor is a leading indicator of journal
influence, and impact factors are routinely used in making decisions ranging
from selecting journal subscriptions to allocating research funding to deciding
tenure cases. Yet journal impact factors have increased gradually over time,
and moreover impact factors vary widely across academic disciplines. Here we
quantify inflation over time and differences across fields in impact factor
scores and determine the sources of these differences. We find that the average
number of citations in reference lists has increased gradually, and this is the
predominant factor responsible for the inflation of impact factor scores over
time. Field-specific variation in the fraction of citations to literature
indexed by Thomson Scientific's Journal Citation Reports is the single greatest
contributor to differences among the impact factors of journals in different
fields. The growth rate of the scientific literature as a whole, and
cross-field differences in net size and growth rate of individual fields, have
had very little influence on impact factor inflation or on cross-field
differences in impact factor.
\\ ( http://arxiv.org/abs/0804.3116 , 995kb)


If I am interpreting this abstract correctly, the authors of this paper seem to be concluding that the differences one sees between fields is predominantly the result of which journals reporting venues (in this case, Thomson Scientific's Journal Citation Reports) choose to index. In other words, they conclude that the dominant factor is basically noise at best and corruptible choices at worst. That's just lovely. (I haven't read the paper, so I am offering no opinion as to whether I agree with the authors' conclusion. I'm just saying that this is an extremely disturbing conclusion.)

How important is somebody?

Here is an interesting-looking article that was just posted on the arxiv. I don't plan on reading it in detail immediately, but its abstract is certainly intriguing. Here's the abstract:


Co-occurrence Network of Reuters News

Arzucan Ozgur, Burak Cetin, Haluk Bingol

Networks describe various complex natural systems including social systems. We investigate the social network of co-occurrence in Reuters-21578 corpus, which consists of news articles that appeared in the Reuters newswire in 1987. People are represented as vertices and two persons are connected if they co-occur in the same article. The network has small-world features with power-law degree distribution. The network is disconnected and the component size distribution has power law characteristics. Community detection on a degree-reduced network provides meaningful communities. An edge-reduced network, which contains only the strong ties has a star topology.
"Importance" of persons are investigated. The network is the situation in 1987. After 20 years, a better judgment on the importance of the people can be done. A number of ranking algorithms, including Citation count, PageRank, are used to assign ranks to vertices. The ranks given by the algorithms are compared against how well a person is represented in Wikipedia. We find up to medium level Spearman's rank correlations. A noteworthy finding is that PageRank consistently performed worse than the other algorithms. We analyze this further and find reasons.

Citation sales curves

I just read the abstract for a paper that discusses 'citation sales curves' as a way of measuring the impact of scientists. I'm not necessarily convinced, but the idea is interesting. Here is the abstract:

\\
Paper: physics/0611284
Date: Wed, 29 Nov 2006 08:24:27 GMT (119kb)

Title: Using time dependent citation rates (sales curves) for comparing
scientific impacts
Authors: Werner Marx, Hermann Schier, Ole Krogh Andersen
Comments: 9 pages, 3 figures, 1 table
Subj-class: Physics and Society
\\
As a simple means for comparing and - if possible - predicting scientific impacts of different researchers working in the same field, we suggest comparing their "sales curves". A sales curve is the number of citations of the researcher's papers per year, the citation rate, considered as a function of time. As examples, we present the citation histories of 10 older well-cited
scientists working in the same field. The sales curve is found to be highly individual, that is, there is a large variation between different scientists' sales curves. For each well-cited scientist, however, the sales curve is steadily increasing as long as he is young and active, and its slope, the citation acceleration, contains the essential information about his impact. The slope averaged over the time of activity of the scientist is roughly independent of time and is a fairly age-independent measure of his scientific impact. In physics and chemistry, well-cited active scientists have time-averaged citation accelerations at the order of 10 citations per year^2 or more. The normal citation acceleration is an order of magnitude smaller. We also show the sales curves for three large research institutes whose sizes have been fairly constant over the last 35 years. These sales curves are quite linear and have slopes at the order of 1 citation per scientist per year^2.
\\ ( http://arXiv.org/abs/physics/0611284 , 119kb)