[soundwave106]

In the above news article, there is a direct link to the magazine with the study. The direct link to the study is here: http://advances.sciencemag.org/content/5/1/eaau4586

The study appeared to correlate two sources to determine the nature of fake news:

A) The primary source was a list of fake news sites compiled by Buzzfeed Media: https://www.buzzfeednews.com/article/craigsilverman/viral-fa...

B) The study was cross-checked with a list of sites from a peer reviewed paper (H. Allcott, M. Gentzkow, Social media and fake news in the 2016 election. J. Econ. Perspect. 31, 211–236 (2017)) and according to the paper was similar.

There is some additional methodology in the study link.

[asabjorn]

Thanks! Buzfeed believes in far-left progressive views and is not known for objective reporting, so using this as a primary source for classifications of fake news would predetermine this outcome.

With this in mind I think we can't trust the studys conclusion.

[fzeroracer]

Why don't you address the actual study instead of disregarding it on a partisan basis?

[asabjorn]

To show that the classification system used as a premise for the study has problems is addressing the methodology of the study, and exactly what I learned in my scientific training.

The fact that the classifications are likely to be politically biased due to them being made by buzzfeed does not make the statement of that fact necessarily political.

[octernion]

old misogynists sharing "wikileaks CONFIRMS hillary sold weapons to ISIS!!!!" being fake news is absolutely a far-left take on reality, hahaha. how are you so bad at this.

[soundwave106]

I do disagree with the lazy "Buzzfeed is left wing therefore your result are automatically invalid" conclusion of the poster. :)

But I do believe that bias could be a genuine concern here.

Both the Buzzfeed article and the Stanford article (https://web.stanford.edu/~gentzkow/research/fakenews.pdf) seems to focus strictly on political fake news.

But Donald Trump fake news is not the only fake news out there.

Given that Donald Trump supporter demographics lean older, I am wondering if an exclusive focus on political fake news (most of which was indeed leaning in the Donald Trump direction) is skewing the results towards the conclusion that people older than 65 share the most fake news.

Would the study look differently if they included sites propagating non-political fake news -- such as celebrity oriented fake news or health woo?

In another case -- Russian disinformation campaigns -- it's well known that they targeted pretty much all sides with divisive Facebook ads on contentious issues or identity politics. Some of the ads certainly are in the "fake news" category (https://www.washingtonpost.com/graphics/2017/business/russia...). The question I would have is whether the same demographics would apply to identity or issues related fake news. I'm not certain here.

It's possible that one can include the above data points and still draw the conclusion; in a tangentially related case, age is one of the factors in other forms of susceptibility to fraud (https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3916958/), so it's possible that age is a factor in not recognizing fake news as easily.

But until then, while the article does make a good case that people older than 65 share the most fake pro-Donald Trump news, I'm not sure I can apply that conclusion to other forms of fake news yet.

[octernion]

i absolutely agree and was taking issue primarily with the lazy political framing of the parent than the broader conclusions of the article as a whole.

i also would like to see a more comprehensive study (esp. including scientific issues, which should actually be easier to measure and less contentious).

[asabjorn]

I would also like a comprehensive study, but this article is not it because it is using classifications of fake news made by tabloid media that does not satisfy any scientific requirement of rigor.

It is ok that buzzfeed is politically biased, but its conclusions and work should also be treated as such with the limitations that presents in the applicability of its conclusions.

[asabjorn]

The point I made was not that Buzzfeed is left-wing as a reason for doubting it as a source for objective classifications, Le Monde is left-wing as well as reputable and likewise is Heterodox academy, but that rigor is needed in your definition of fake news and that the classification should be statistically controlled for political bias. Buzzfeed seems like a rather odd source if you seek to achieve those goals.