Wikipediocracy post analyzing favored Wikipedia sourcing

General discussion about Wikipedia and other Wikimedia projects
Post Reply
User avatar
The Devil's Advocate
Posts: 357
Joined: Wed Aug 23, 2017 7:50 pm

Wikipediocracy post analyzing favored Wikipedia sourcing

Post by The Devil's Advocate » Sun Aug 26, 2018 9:26 pm

Sashi put out a blog post on WO analyzing sources on Wikipedia by crawling through internal search results, specifically focusing on the occurrence of links in the source code to get best results. The best amount of effort put into this is pretty exceptional and it is probably the only decent blog post they've had up in a year. One flaw I would note is with the implication this search would be useful to indicate the use of social media as references on Wikipedia. In fact, Wikipedia articles often have links to official social media accounts of article subjects and in some cases, such as Joe Biden and Donald Trump, the only direct links to Twitter are to point to their official accounts on those sites.

Despite including 578 sources, it is still not very comprehensive as it excludes many of the major Chinese news sources such as Xinhua, China Daily, and CCTV. NBC is also not fully represented as sometimes it is cited as "today.com" and that was not include in his list of searches. BBC is slightly underrepresented as they also have .com address with nearly 25,000 results, most of which do not appear to be duplicative. Newscorp's main Australian outlet news.com.au is also excluded. All that said, I still think this is a pretty useful analysis as well as the accompanying datasheet. A way to refine searches and trim down non-reference links is by including details of the citation templates. For instance "url=https://twitter.com" returns less than 20,000 results compared to the 35,000 obtained by Sashi's search, though that still includes some instances of just noting official Twitter links.

Even with these imperfections, it is a pretty good dataset to examine for many reasons. Notable that the Guardian appears to be so highly-cited, being linked more than CNN or The Washington Post, given it is a far more biased source than the latter two outlets. Particularly interesting given that there are six news outlets in the top 25 of his list and only one, The Daily Telegraph, is known for a conservative slant. All of the others are typically understood to have a left-of-center bias in varying degrees. It is possible there is a non-political explanation for some of that favoritism, but it is hard to overlook the fact the Guardian seems to be frequently favored by Wikipedians beyond just in contributions to articles. Wales himself sat on the board at one time, after all.

sashi
Posts: 116
Joined: Mon Feb 27, 2017 6:26 pm

Re: Wikipediocracy post analyzing favored Wikipedia sourcing

Post by sashi » Mon Aug 27, 2018 12:05 am

Thanks for this review, TDA. I'm going to look into some of the missing culprits (right now I'm working exclusively on the sorted version, though it can be re-transformed into a longlist again... with a little work). I'd worked on Mexico & Brazil a bit in the previous version, but now have added all of your suggestions.

Concerning twitter, the problem I see is that I lose those references that are just in <ref> tags if I specify url= and there is a difference of a couple thousand already between twitter.com and sptth://twitter.com (e.g. §). Also, social media / self-published is a class that is a lot bigger than just twitter. I just did a sum and came up with close to half a million pointers! (>2x the NYT :twisted: ) (I'd forgotten tripod... :lol: )

and National Geographic. :o

I've updated BBC... and news.com.au, NBC / today.com as well as the three Chinese sites you've mentioned. Thanks a lot for the many tips! It's not a finished project. ^^

User avatar
Auggie
Posts: 390
Joined: Sat Feb 25, 2017 2:00 pm

Re: Wikipediocracy post analyzing favored Wikipedia sourcing

Post by Auggie » Tue Aug 28, 2018 11:32 pm

Come on Sashi. What about us? :mrgreen:
no love from Sashi wrote:Acknowledgements
I would like to thank the members of the Wikipediocracy forum, the Wikipedia Sucks (and so do its critics) forum, and the Gender Desk blog for either encouraging me in this enterprise, offering suggestions on earlier drafts, or for including sources in their posts that ended up being among those listed in this study. Finally, I would like to close by acknowledging that, concerning methods, I have left much of the work to the imagination of the reader.

sashi
Posts: 116
Joined: Mon Feb 27, 2017 6:26 pm

Re: Wikipediocracy post analyzing favored Wikipedia sourcing

Post by sashi » Wed Aug 29, 2018 6:54 pm

Yeah, I should have, especially since the elephant analogy was born here.
love from sashi wrote:Thanks, Auggie.
and speaking of sourcing... I would be interested in what you think of these articles, especially the one about the Dark Side. ^^

Renée Bagslint
Posts: 298
Joined: Sat Nov 25, 2017 1:43 pm

Re: Wikipediocracy post analyzing favored Wikipedia sourcing

Post by Renée Bagslint » Wed Aug 29, 2018 9:29 pm

Getting some traction on the wikimedia-l list. I liked this one
Robert Fernandez wrote:Interesting metrics and ideas here, but nobody's going to take your research particularly seriously if you choose to post it in an open sewer.

Proabivouac
Posts: 370
Joined: Thu Aug 24, 2017 7:01 pm

Re: Wikipediocracy post analyzing favored Wikipedia sourcing

Post by Proabivouac » Thu Aug 30, 2018 1:34 am

Renée Bagslint wrote:
Wed Aug 29, 2018 9:29 pm
Getting some traction on the wikimedia-l list. I liked this one
Robert Fernandez wrote:Interesting metrics and ideas here, but nobody's going to take your research particularly seriously if you choose to post it in an open sewer.
As opposed to Robert "Gamaliel" Fernandez' now-moribund Signpost, with its joke columns about penis size?

User avatar
Auggie
Posts: 390
Joined: Sat Feb 25, 2017 2:00 pm

Re: Wikipediocracy post analyzing favored Wikipedia sourcing

Post by Auggie » Thu Aug 30, 2018 2:11 pm

sashi wrote:
Wed Aug 29, 2018 6:54 pm
Yeah, I should have, especially since the elephant analogy was born here.
love from sashi wrote:Thanks, Auggie.
and speaking of sourcing... I would be interested in what you think of these articles, especially the one about the Dark Side. ^^
Cool thanks. There is a lot of stuff there. I will check it out.

Post Reply

Who is online

Users browsing this forum: No registered users and 3 guests