On Thu, Jul 11, 2024 at 11:09 AM J. Nathan Matias natematias@gmail.com wrote:
Thanks for sharing, and many thanks to Tilman Bayer for assembling this summary.
Thanks Nathan!
I'm really intrigued by this study about differences in sentiment and language, especially because work on moral foundations theory https://en.wikipedia.org/wiki/Moral_foundations_theory in political science has found that different political groups in the US are associated with different approaches to care/harm, fairness/cheating, loyalty/betrayal, authority/subversion, sanctity/degradation, and liberty/oppression.
A hypothesis: it's possible that these differences, if they exist, are the result of choices by conservatives and liberals in what they choose to do and how they frame their own work. For example, a politician that built their campaigns around outrage and anger might be expected to have more negative material on their Wikipedia page, while someone who built their career around constructive visions of society might have quite different language on their page.
Yes, that's a great point. However (not sure if you saw it mentioned in the review), Rozado actually already tried to address a similar concern, see the paper's "Discussion" section. Not regarding Haidt et al.'s moral foundations theory in general, but about some more specific - but disputed - research results in favor of what he calls the "self-sorting emotional affect hypothesis", which "suggested that conservatives themselves are more prone to negative emotions and more sensitive to threats than liberals." He offers some valid counterarguments in that case.
But in general I think you're right: Even if research demonstrates the existence of glaring disparities on Wikipedia that appeal to an intuitive sense of fairness (as Rozado's findings have evidently done successfully for many people, see also the media coverage mentioned at the end of the review), it's worth being skeptical of claims that these are *caused* by Wikipedia's own biases. And apart from the single remark mentioned above, the paper indeed does not do a lot to explore the possibility that these disparities could be explained by factors outside of Wikipedia. See also the asymmetric polarization thesis that I mentioned in the review, which has received a lot of attention in (at least US) political science and would seem quite relevant here too.
All that said, to be fair, the paper shares these shortcomings with quite a few other papers about gaps and biases on Wikipedia, and still seems more circumspect overall than some past publications that received substantial media attention and enthusiasm despite a deeply flawed methodology. Of course Wikipedians will often be inclined to commit the opposite fallacy, dismissing uncomfortable findings too easily. But that should not stop us from calling out unsubstantiated bias claims and shoddy research methods. As mentioned at the beginning of the review, the Wikimedia Foundation, to its credit, did exactly that some months ago, in no uncertain terms, in case of a different publication that had also claimed to demonstrate (a different form of) political bias on English Wikipedia, and had likewise been commissioned by an advocacy organization and authored by a researcher with relevant academic credentials (having previously published peer-reviewed research about gender bias on Wikipedia).
Regards, Tilman
All the best,
--Nathan
On Thu, Jul 11, 2024 at 4:44 AM ABDUL-RAFIU FUSEINI < abdulrafiu.fuseini264@gmail.com> wrote:
The June 2024 issue of the Wikimedia Research Newsletter is out: https://meta.wikimedia.org/wiki/Research:Newsletter/2024/June
In this issue : Report by conservative think-tank presents ample quantitative evidence
for
"mild to moderate" "left-leaning bias" on Wikipedia. Briefly
*** 8 recent publications *** were reviewed or covered in this issue.
Alhaji Darajaati on behalf of the Newsletter team
Wikimedia Research Newsletter
- https://meta.wikimedia.org/wiki/Research:Newsletter/
- Follow us on Twitter/X: https://x.com/WikiResearch
- Follow us on Mastodon: https://mastodon.social/@wikiresearch
- Like us on Facebook: Facebook.com/WikiResearch/
- Receive this newsletter by mail: Research-newsletter Mailing List
Wiki-research-l mailing list -- wiki-research-l@lists.wikimedia.org To unsubscribe send an email to
wiki-research-l-leave@lists.wikimedia.org
-- J. Nathan Matias http://natematias.com/ : Cornell University : Citizens and Technology Lab https://citizensandtech.org : Coalition for Independent Technology Research http://independenttechresearch.org/ : social.coop/@natematias : blogs https://natematias.com/external-posts/ : daylight time photos https://social.coop/tags/outdoormoments _______________________________________________ Wiki-research-l mailing list -- wiki-research-l@lists.wikimedia.org To unsubscribe send an email to wiki-research-l-leave@lists.wikimedia.org