Mark’s first post on Tim Groseclose’s liberal-media-bias book (which he linked to in this later post and this one) links, if one clicks on the phrase “grossly tendentious,” to a critique of Tim’s book by Brendan Nyhan in Perspectives on Politics (Vol. 10, No. 3 [September 2012], 767-771).
That issue of Perspectives actually contains a whole forum on the book, and readers might be interested in the other contributions. So, as an easy public service, here are links to
Nyhan’s piece in .html (and .pdf);
The contribution by Nolan McCarty in .html (and .pdf);
The contribution by Justin H. Gross, Cosma Rohilla Shalizi and Andrew Gelman in .html (and .pdf);
The contribution by Nancy L. Rosenblum in .html (and .pdf);
The contribution by Kathleen Hall Jamieson in .html (and .pdf).
Alas (for current readers, though necessary for the long-term existence of high-quality academic journals), I think all of these are behind an academic paywall.
On this question I am both unbiased and unhelpful due to ignorance. Though I know Tim slightly and like and trust him as a person—while taking great exception to his politics—I haven’t gotten around to reading either the book or the critiques (some of whose authors I also know slightly, or better than that). But as possibly the world’s last surviving Millian, I have to think that people interested in this stuff, including me, eventually, will benefit from a variety of contributions by highly regarded political scientists.
Moreover, at a quick glance, many of those contributions contain extensive citations to further research and commentary on the book, most of it scholarly and in some cases quite technical. When it comes to political science, as well as policy, there’s sometimes no substitute for an expert take.
Thanks!
I would *really* like to read the Gross, Shalizi, and Gelman piece.
If the reason you can’t read it is the academic paywall, someone might hypothetically be very willing to send you a copy if you emailed him..
We have of course a non-paywalled version of our piece:http://www.stat.cmu.edu/~cshalizi/leftturn/
Whether the other contributors have un-gated versions I couldn’t say. (I am also not able to stick around for this discussion.)
Thanks Cosma! I wouldn’t think to presume about the timesink of a blog discussion but it’s important to note that my better half has been wondering if things were ok, due to the somewhat attenuated rate of mind candy reviews… 🙂
Ok, I was going to summarize Gross, et al., but after digesting it I believe if you’re interested you should just read it.
The piece is a dense sequential litany of the modelling errors and omissions of Groseclose, expressed in the nicest possible way.
I’m not going to be nice here. The practical, intended effect (essentially self-admitted) of Groseclose’s “work” is to cloak pure propaganda in the language and stylisms of social science. The implicit goal is to accumulate those 200+ and counting citations on Google Scholar.
I’ve really got to wonder about the state of a “science” which legitimates such stuff. Economics suffers from nearly identical meta problems as well, cf R&R.
Let’s get meta in an orthogonal direction… As an amateur reader I was amused that one outcome of professional editing was the propagation of a superfluous “to” (p. 777, 7 lines from end). A very few phrasings were arguably made more felicitous. But the value “add” seems to mostly amount to a stripping off of the really interesting stuff, which lives in the online appendices.
oops: “to” -> “of”. Memory spoke, and it lied.
Yes, the Gross et. al. piece is one of the most devastating reviews I’ve ever read - all the while remaining unremittingly professional and positive.
¨Alas (for current readers, though necessary for the long-term existence of high-quality academic journals), I think all of these are behind an academic paywall.¨
The real work in academic publishing - the review process - is done unpaid by academics and covered by their salaries. The research published is typically paid for out of public or general university funds, never by the publishing house. Authors format their own work, so typesetting has gone. So why are paywalls necessary for any learned journals?
The specific and rather menial services provided by the publishing house, the website, archiving, physical printing and distribution, and the like, are low in cost and easily lost in the budget of a university library. It´s not clear that physical printing and distribution, the most expensive item, any longer adds value. If it does, it can be hired in India or Mississippi or another Third World country and charged to the few dinosaur users at cost.
Profit-making academic publishing is an obsolete parasitic rent-seeking nonsense. No self-respecting academic should publish anywhere that imposes a paywall on scholars - or - I would add, the ordinary citizen who foots the bill. Aggregating sites like Jstor should move as quickly as possible to free-to-view. See, for instance, Cambridge mathematics professor Tim Gowers´ boycott of Elsevier, and the NSF open access policy.
That is a damned good question, and one the publishers have yet to answer satisfactorily. We do the work, we do the reviewing and editorial work, we pay them to publish the work (page charges) and they then charge our libraries amounts that range from ridiculous to extortionate to unconscionable for the right to allow us to access our work.
The cost to set up and maintain a small server farm is negligible: every university with a library could afford to do it for less than the costs of subscriptions to one of their high-cost departments (say, Biology).
While it’s true that peer reviewers and authors get paid zero, the editors of most top journals (the ones that are the hardest to get into and therefore by definition get the most submissions, and involve the most work) are paid nontrivial amounts-and have to be, or else doing the job would be even more irrational than it currently is. Someone at the press still has to do the formatting, proofreading, and so on that make journal articles easier to read than printouts from blog posts.
In other words, journal publishing, like all publishing, requires skilled work that must be paid for. I agree that the current cost structure is a scam and that the cost for non-academics to download and print each article could be drastically reduced-but not to zero. Free content is, as usual, worth what you pay for it. That was in fact the thesis of my very first blog post:
https://thesamefacts.com/2006/01/msm-mainstream-media/why-news-should-cost-a-lot/
¨Someone at the press still has to do the formatting, proofreading, and so on …¨ I´m genuinely curious. Do they really do this still? In which fields? In Gowers` field of number theory, you are not going to find a printer who understands the material at all, so perhaps the mathematicians prefer to format their proofs themselves.
If journal reviewers are paid, which is interesting news to me,[Sorry, careless reading!] The sensible way of funding the cost of editors is by submission fees covered in the research grant. It is not efficient to charge end-users, since the marginal cost of downloads is zero (strictly speaking it´s just too cheap to meter.)You don´t address the democratic argument that the citizens have paid for the work already through taxes, or in a few cases tax breaks to the endowments of private universities. We are entitled to see the work product without paying twice. The one fair criticism of the English ¨Climategate¨ researchers was that they did not think they had a duty to provide their data freely to malicious denialist hacks. The reviews conceded that scientists do have such a duty of transparency. Science is not a closed shop.
Submission fees paid for from grants are fine for grant-funded research. That leaves those of us who don’t get funding rather out in the cold.
Much more sensible to have the journal publishers directly publicly funded or paid for by a consortium of the big universities whose library subscription fees provide most of the revenue.
Still, I’m rather glad that Berkeley Economic Press thought it might be able to makes some money publishing the Journal of Drug Policy Analysis (since acquired by de Gruyter). No agency, and no foundation, was going to be interested in supporting that journal.
They don’t in my field (Statistics). At most, they take Word or LaTex code and lightly reformat it, as in place it in columnar format (if that’s what the journal does) and put the tables and graphics in the right places. Honestly, I don’t know why they don’t have us doing that in our submissions.
“The one fair criticism of the English ¨Climategate¨ researchers was that they did not think they had a duty to provide their data freely to malicious denialist hacks. ”
IIRC, most of that was in response to requests for information which was purchased from government offices, under conditions of no further distribution. In which case the denialists can ask ExxonMobil for the money to buy it from the original source.
Some of the rest was that the denalists would do ‘denial of service’ attacks, such as asking for data for one city for one year, then another city for the same year, ….., then the original city for another year,……
How much genuinely improper withholding of data was found to have been done? I
Not always; in the case of Elsevier, it’s likely that Merck paid for the first three steps.
I promise you that you are not the world’s last surviving Millian. Among a million, I rush to assure you of our solidarity (not to be confused with unanimity as the solidarity of those who vigorously debate how to achieve shared goals is the most solid, fruitful, sweet, shining, flaming and uh mixed metaphors). I stress that 999,999 of us share no blame for the horrible pun.
I can’t read these articles and I’m not really willing to spend at least $20 just to participate more fully as a commenter at the RBC. The thing that I don’t understand and can’t find is a succinct description of the methodology. Without such a summary, I think talking about the study (or even debating whether it’s a study or the unmoored ravings of a true believer) seems impossible.
Can anyone with knowledge about the “study” we’re discussing describe its methodology?
1. Assign every think tank a liberal/conservative score based on the voting records of Members of Congress who cite its work.
2. Assign every news outlet a score based on the scores of the think-tanks it cites.
That is: call an outlet “neutral” if its citation sources match those of the median-voting Congressman, and “biased” otherwise.
This turns out to be mathematically equivalent to the following approach:
1. Assume liberal media bias.
2. Randomly try methods in order of increasing silliness.
3. Stop when you find one that produces your assumption as a result.
Mark,
if your description is even remotely accurate, this “study” is nothing more than agitprop that would receive a failing grade in any college course not taught by Rush Limbaugh.
Again, if that’s the methodology, I find it astonishing that a reputable academic journal would publish it or that a group of distinguished academics would actually take it seriously enough to devote an entire issue of a journal to such obvious slop.
The hoax Sokal paper must have got a stratospheric number of citations. And William McGonagall is widely quoted.
Hi all - here is an ungated copy of my piece.
Thanks! And thanks to Cosma Shalizi too.
Now I’m feeling awful for not linking to the ungated versions. I guess I still can’t quite believe that journals don’t sue their own authors for providing such.
I know I said I don’t have time for a blog discussion now, and I don’t, but I am astonished that a scholar of Andrew’s caliber feels commercial journal publishing is anything other than a racket at the public expense. For our paper, the paid staff of the journal did not organize the session, did not format the manuscripts for publication (we did), did not apply a uniform style to the references (though they told us that our style didn’t conform to the journal’s), and did minimal copy-editing. For this, Cambridge University Press, an ostensibly non-profit entity, purports to charge $30/article; of course the point of such prices is to force university libraries (overwhelmingly taxpayer-supported) to subscribe to the journal. Experience in fields like statistics, computer science and physics shows that the real costs of running a journal in the present day and age can be driven far below what many might imagine (cf. “An Efficient Journal”, and especially the back-and-forth in the comments where a professional editor simply refuses to believe the well-established facts about the premier journal in machine learning). It wouldn’t surprise me to learn someone like Elsevier had contemplated suing authors who make their papers freely available, but that’s because they’re obsolete parasites whose profit comes entirely from hindering the dissemination of knowledge. A better system has been technically feasible for about twenty years now.