Wednesday, 3 May 2017

Some papers just shouldn't be published

I don't post about all of the papers I read. Some of them are just less interesting than I thought when I first added them to the must-get-around-to-reading-that-sometime pile. Rarely though, there is a paper that is just pretty awful. This one by Vsevolod Andreev (Chuvash State University in Russia) is one example, entitled "Will there be a revolution in Russia in 2017?" and published in the Journal of Policy Modeling (ungated here) in 2015 (it's been in the pile for a while, but now seemed like a good time to crack it open).

The paper mostly appears mostly to be an egregious attempt to inflate the author's own citation counts - 11 of the 29 papers cited in this article are the author's own work. Take this paragraph for example:
Also in my previous studies (Andreev & Jarmulina, 2009; Andreev & Karpova, 2007; Andreev & Karpova, 2008; Andreev & Semenov, 2010a; Andreev & Semenov, 2010b; Andreev & Semenov, 2010c; Andreev & Semenov, 2012; Andreev & Semenov, 2013; Andreev & Vasileva, 2009; Karpova & Andreev, 2007; Karpova & Andreev, 2008) the mathematical models of dynamics of socio-economic systems, created on the principles of predator-prey models, are proposed and investigated. These models were applied for analysis at different time stages of dynamics of socio-economic systems of Russia (Andreev & Jarmulina, 2009; Andreev & Karpova, 2007; Andreev & Karpova, 2008; Andreev & Semenov, 2013; Andreev & Vasileva, 2009; Karpova & Andreev, 2007; Karpova & Andreev, 2008) and of the USA (Andreev & Semenov, 2010a; Andreev & Semenov, 2010b; Andreev & Semenov, 2010c; Andreev & Semenov, 2012). These studies results quite adequately describe the observed real situation.
That's right. That paragraph includes 22 citations of the author's own work, which might be very impressive if it was a solid body of research in a neglected subfield. But in this case, none of those other papers has been published in a reputable journal, and I wouldn't be at all surprised if they weren't just very similar versions of this paper (given the titles in the reference list). Now, I do self-cite in my own papers where appropriate, as does every other researcher, but that paragraph is really taking the piss.

Overall, the paper would be hilarious, if it hadn't actually been accepted for publication. Of course, if it had been published in Economic Inquiry, I might have put it down to being a bit of a joke (Economic Inquiry is the journal responsible for classics like "Riccardo Trezzi is immortal" (see my post on that one), and "On the Efficiency of AC/DC: Bon Scott versus Brian Johnson"). However, in this case Andreev's paper is mostly gibberish. He uses a predator-prey model (which sounds impressive, but it's really not) involving relationships between incomes, GDP, government spending on basic and applied research, population and the outflow of capital. The model appears to perform reasonably well in-sample, but extrapolating forward in time it all goes a bit haywire, leading Andreev to note that:
The blow-up regime is understood to occur when the behavior of one or several functions xi(t) of the system state begin to grow uncontrollably during the small time interval...
And then conclude that:
...it is possible to conclude that in Russia there are symptoms of a revolutionary situation at the end of 2016 and at the beginning of 2017.
Because the dodgy model extrapolated out-of-sample exhibits weird dynamic behaviour (which isn't at all surprising for a nonlinear model), Russia will have a revolution in 2017. I guess we will see. Peer review is supposed to weed out this sort of rubbish. As a reviewer, I would have rejected it in a heartbeat.

No comments:

Post a comment