This document summarizes research on fake news and disinformation. It discusses how fake news spreads online, often through social media sites like Facebook, and how a small proportion of people consume most fake news. While few people directly consume fake news, its presence online can still influence beliefs and undermine trust in real news. The document also examines how fake news creators target people's emotions for profit or propaganda, and how emotions spread online. It concludes that addressing fake news will require a multi-stakeholder approach that includes education, changes to digital platforms and advertising, and efforts to increase media literacy.
1. FAKE NEWS &
THE ECONOMY OF EMOTIONS:
PROBLEMS, CAUSES, SOLUTIONS
Professor Vian Bakir
Bangor University, Wales, UK
Desinformation vs demokratin – vad kan vi göra?
Kalmar, Sweden 20-22 November
2. Trump vs. Clinton 2016
Deception, voter profiling, targeting online
Fake News and The
Economy of Emotions:
Problems, causes, solutions
Vian Bakir & Andrew McStay
Digital Journalism Vol 6(2), 2018, pp.
154-175
6. • 23 oral evidence sessions
• >170 written submissions
• Evidence from 73 witnesses
• Asked >4,350 questions at hearings
Our submissions:
• 2017. Summary And Analysis Of All Written
Submissions On How To Combat Fake News
(Up To April 2017)
• 2017. Fake News: Media Economics and
Emotional Button-Pushing
• 2017. Fake News: A Framework for Detecting
and Avoiding Propaganda
Inquiry’s reports:
• Digital, Culture, Media and Sport Committee. 2018.
Disinformation and ‘fake news’: Interim Report.
Jul. House of Commons 363.
• Digital, Culture, Media and Sport Committee. 2019.
Disinformation and ‘fake news’: Final Report. 18
Feb House of Commons 1791.
7. 960,000 Facebook shares, likes & comments
What is fake news?
(1) Low in facticity
(2) Intention to deceive
(3) Looks like genuine news story
(Egelhofer & Lecheler 2019)
8. What does fake news look like?
• Shorter & less informative than
genuine news
• Uses less complex, more
personal language
• Longer titles which contain
main claim of article
(Horne & Adal 2017)
- Targeted for audiences who aren’t likely to read beyond titles
- Persuasion relies on heuristics e.g. conformance of the info to one’s beliefs
9. HOW MUCH
NEWS ONLINE IS
FAKE?
Study 1 Buzzfeed, 6 hyper-partisan
Facebook pages - 7 days, Sep 2016
• Finds mix of true & false or mostly false:
• 38% of posts in 3 right-wing pages
• 19% of posts from 3 left-wing pages
• Falsehoods are often claims/ accusations
against people, companies, police,
movements e.g. Black Lives Matter,
Muslims, "liberals”, "conservatives"
• Drives division & polarisation
(Silverman, Strapagiel, Shaban & Hall 2016)
10. HOW MUCH
NEWS ONLINE IS
FAKE?
Study 2 Computational analysis of
Global Database of Events, Language &
Tone (2014-16)
• Finds:
• 60 fake news websites
• 171,365 stories from fake news sites
• Content from fake news sites is rising
• Fake news is responsive to agendas of
partisan media on economy, education,
environment, international relations,
religion, taxes, unemployment.
(Vargo, Guo & Amazeen 2018)
12. HOW MANY
PEOPLE
CONSUME FAKE
NEWS?
• Small, concentrated
proportion of people
• Deep ‘echo chamber’
Study 1: US national survey + individual web traffic histories
(Oct - Nov 2016) (Guess et al. 2018)
• 1 in 4 Americans visited a fake news website
• Mostly Trump supporters & via Facebook
• Fact-checks of fake news almost never reach its consumers.
• 6 in 10 visits to fake news sites are from 10% of people with
most conservative online information diets
Study 2: online visitation data months before & after 2016
US presidential election (Nelson & Taneja 2018)
• Fake news audience is small, disloyal group of heavy Internet
users
Study 3: 300 ‘false news’ websites in each of France &
Italy measuring reach, attention, no. of interactions on
Facebook (Fletcher, Cornia, Graves & Nielsen 2018)
• Most reach <1% online population, far less than popular
news sites:
- Le Figaro (France) average monthly reach 22.3%
- La Repubblica (Italy) average monthly reach 50.9%
13. NOT MANY PEOPLE CONSUME FAKE NEWS,
SO DOES FAKE NEWS REALLY MATTER?
Exposure to fake news may be low, but it is corrosive
• We are bad at detecting deception
• Presence of fake news in ecosystem alters what we believe to be true & what
we trust
• False information spreads rapidly via social media
• Digital news is increasingly emotionalised, fake news is negatively emotive, &
emotions are contagious online
Yes!
14. NOT MANY PEOPLE
CONSUME FAKE NEWS,
SO DOES FAKE NEWS
REALLY MATTER?
Study 1. US national online
survey of 3,015 adults (Nov –
Dec 2016) (Silverman & Singer-Vine
2016)
• fake news headlines fool US adults
75% of the time
Study 2 of 7,800 US middle
school, high school & college
students’ ability to assess online
information sources (Stanford
History Education Group 2016)
• they ‘are easily duped’
YES!
We are bad at
detecting
deception
15. NOT MANY PEOPLE
CONSUME FAKE NEWS,
SO DOES FAKE NEWS
REALLY MATTER?
Study 1. Repeated exposure to fake news
headlines increases their perceived accuracy
(Pennycook, Cannon & Rand 2018).
• despite low level of overall believability
• even when fact checkers label stories as
‘contested’
• even when stories are inconsistent with reader’s
political ideology
Study 2. Exposure to elite discourse on fake news
(US) leads to less trust in media & less accurate
identification of real news (Van Duyn & Collier
2019).
Study 3. A fake news attack by Trump significantly
reduces perceived media accuracy & media trust
among Trump supporters (Guess et al. 2017).
YES!
Presence of fake
news in ecosystem
alters what we
believe to be true &
what we trust
16. NOT MANY PEOPLE
CONSUME FAKE NEWS,
SO DOES FAKE NEWS
REALLY MATTER?
Study 1 - fake news consumption up
to & following 2016 US presidential
election (Nelson & Taneja 2018)
• visits to fake news sites originate
from social network sites at much
higher rate than visits to real news
sites
Study 2 - differential diffusion of all
verified true & false news stories on
Twitter (2006 –17) (Vosoughi et al. 2018)
• Falsehood – esp. false political news -
diffused significantly farther, faster,
deeper & more broadly than truth.
• False news spreads more than truth because
humans (not bots) are more likely to
spread it.
YES!
False information
spreads rapidly via
social media
17. NOT MANY PEOPLE
CONSUME FAKE NEWS,
SO DOES FAKE NEWS
REALLY MATTER?
Study 1. Digital news is increasingly
emotionalised (Beckett and Deuze 2016)
• mobile digital media - increasingly personalised
(algorithmically formed filter bubbles)
• & intimate ( ‘always on’ hand held smart devices;
personal & public networks interconnect)
• uses emotion to virally engage readers in
increasingly competitive news ecology (e.g.
clickbait)
Study 2. of US political blogs, talk radio &
cable news analysis programs (Sobieraj & Berry
2011)
• finds extensive ‘outrage discourse’ (to provoke
visceral response via anger, fear, moral
righteousness)
• thrives in narrowcasting environment where
networks reach out to smaller, homogeneous
audiences & can afford to offend
Study 3 - differential diffusion of all verified true &
false news stories on Twitter (2006 –17) (Vosoughi
et al. 2018)
• false stories inspired fear, disgust & surprise in
replies
• true stories inspired anticipation, sadness, joy, trust.
YES!
Digital news is
increasingly
emotionalised,
fake news is negatively
emotive …
18. NOT MANY PEOPLE
CONSUME FAKE NEWS,
SO DOES FAKE NEWS
REALLY MATTER?
Study 1. secretly optimised 689,003 people’s
News Feeds to see how emotions expressed
by others on Facebook influence our own
emotions (Kramer et al. 2014)
• finds emotions are contagious on
Facebook
• i.e. demonstrates ability to algorithmically
sort and manipulate online ‘fellow
feeling’
Study 2 of >165,000 tweets during 2011
German elections (Stieglitz & Dang-Xuan
2013)
• finds emotionally charged tweets
retweeted more often & quickly than
neutral ones
• heavy Twitter users post more
emotionally charged tweets
Study 3 of 34,770,790 English-language
tweets in 2010 (Thelwall, Buckley & Paltoglou
2011)
• finds popular events are normally associated
with increases in negative tweets
YES!
Digital news is increasingly
emotionalised,
fake news is negatively emotive
…
& emotions are
contagious online
19. TARGETING OUR EMOTIONS
• Citizens are exposed to fake news & false information via
behaviourally targeted adverts & posts
• Targeting is emotionally loaded for commercial &
propagandistic gain
20. “… these companies - Google,
Facebook and YouTube - should be
seen as … artificial intelligence
systems. The fuel for AI systems
is information that lets you do
better prediction…. knowing
to predict what you will be
vulnerable to, what messages
you will be influenced by or
invulnerable to, and what
advertisers might want to target
you”
Tristan Harris, Co-founder
& Exec Dir., Center for
Humane Technology
Oral Testimony,
UK Fake News Inquiry,
HC 363. 2018 22 May
21.
22.
23. TARGETING OUR EMOTIONS
• Upsurge in fake news sites on Facebook
during 2016 US presidential election
• by computer science students & teenagers in
Macedonia
• launched many US politics websites e.g.
USADailyPolitics.com
WorldPoliticus.com
DonaldTrumpNews.co
• Fake news site earns income by:
• attracting visitors & serving them
behaviourally targeted ads (paid by ad
network)
• from click-throughs on ads
• Experiments with left-leaning content under-
performed compared to pro-Trump content on
Facebook
(Mustafaraj & Metaxas 2017, Allcott & Gentzkow 2017)
for commercial &
propagandistic gain
24. TARGETING OUR EMOTIONS
Leave.EU – funded by Arron Banks
• ‘My experience of social media is it is a
firestorm that, just like a brush fire, it blows over
the thing. Our skill was creating bush fires and
then putting a big fan on and making the fan
blow’ (Banks 2018)
• immigration set ‘the wild fires burning’
Leave.EU was behind a fake undercover
investigation video during Brexit campaign that
went viral on Facebook (Channel 4 News 2019)
• Claimed to show how easy it is to smuggle
migrants to UK across English Channel.for commercial &
propagandistic gain
27. HORIZON LINE
AUTOMATED FAKE NEWS CREATION USING AI
• Deepfakes (real people with faces
swapped) - video/audio + AI
• Mal-uses of AI-generated synthetic media and
deepfakes
• Deepfakes explained, USA today
• Algo-journalism (AI writes simple
stories) + emotional AI (sentiment
analysis) to “feel into” users/groups at
scale
28. PUSHING PEOPLE’S
EMOTIONAL BUTTONS
DECEPTION IN PRIVATE
(VIA TARGETED FAKE
NEWS & ‘DARK POSTS’) … … produces
polarised, angry
society that
disbelieves facts &
has no shared basis
of reality
30. DISINFORMATION
MEDIA ECOLOGY
HAS MANY
STAKEHOLDERS
• Education – increase media literacy, emotional
literacy
• Media organisations – promote fact-checking,
pluralistic media ecology
• Digital intermediaries eg Google, Facebook -
downgrade & label fake news
• Advertisers – don’t fund fake news sites
• Professional persuaders & PR – be less deceptive
• Intelligence agencies – detect/block fake news
31. Solutions ?
1. Education
– media literacy
2. Media organisations
3. Digital intermediaries e.g.
Google, Facebook
4. Advertisers
5. Professional persuaders/ PR
6. Intelligence agencies
BUT
… hard to recognise fake news,
confirmation bias
… journalism struggles financially
… secret algorithms, business model of
attention economy
… money rules!
--- deception, bad actors, censorship fears
--- secrecy, info warfare arms race
EACH
SOLUTION
HAS
PROBLEMS
32. REFERENCES
• Allcott, H., & Gentzkow, M. 2017. Social media and fake news in the 2016 election.
Journal of Economic Perspectives, 31(2): 211–236.
https://www.aeaweb.org/articles?id=10.1257/jep.31.2.211
• Bakir, V. & McStay, A. 2018. Fake News and the Economy of Emotions: Problems,
Causes, Solutions. Digital Journalism, 1-22.
• Beckett. C. & Deuze, M. 2016. On the role of emotion in the future of journalism. Social
Media + Society 2(3): 1–6.
• Egelhofer, J. L. & Lecheler, S. 2019. Fake news as a two- dimensional phenomenon: a
framework and research agenda, Annals of the International Communication
Association, 43:2, 97-116.
• Guess, A., Nyhan, B. & Reifler, J. 2018. Selective exposure to misinformation: Evidence
from the consumption of fake news during the 2016 U. S. presidential campaign.
https://www.dartmouth.edu/~nyhan/fake-news-2016.pdf
• Guess, A., Nyhan, B. & Reifler, J. 2017. “You’re fake news!” The 2017 Poynter media
trust survey.
https://poyntercdn.blob.core.windows.net/files/PoynterMediaTrustSurvey2017.pdf
• Horne, B. D., & Adal, S. 2017. This just in: Fake news packs a lot in title, uses simpler,
repetitive content in text body, more similar to satire than real news.
https://arxiv.org/abs/1703.09398
33. REFERENCES
• Kramer, A.D.I., Guillory, J.E. and Hancock, J.T. 2014. Experimental evidence of massive-scale
emotional contagion through social networks, Proceedings of the National Academy of
Sciences, 111 (29): 8788–90.
• Mustafaraj, E. & Metaxas, P. T. 2017. The fake news spreading plague: Was it preventable?
http://arxiv.org/ abs/1703.06988
• Nelson, J. L. & Taneja, H. 2018. The small, disloyal fake news audience: The role of audience
availability in fake news consumption. New Media & Society, 20(10), 3720–3737.
• Pennycook, G, Cannon, T. D. & Rand, D.G. 2018. Prior exposure increases perceived accuracy
of fake news. Journal of Experimental Psychology: General, 147(12), 1865-1880.
• Silverman, C., Strapagiel, L., Shaban, H. & Hall, E. 2016. Hyperpartisan Facebook pages are
publishing false and misleading information at an alarming rate. Buzzfeed News, 20 October.
https://www.buzzfeed.com/craigsilverman/partisan-fb-pages-analysis
• Silverman, C. & Singer-Vine, J. 2016. Most Americans who see fake news believe it, new
survey says. Buzzfeed, 6 December.
https://www.buzzfeednews.com/article/craigsilverman/fake-news-survey
• Sobieraj, S., & Berry, J. M. 2011. From incivility to outrage: Political discourse in blogs, talk
radio, and cable news. Political Communication, 28(1), 19–41.
34. REFERENCES
• Stanford History Education Group. 2016. Evaluating Information: The Cornerstone of Civic
Online Reasoning.
https://sheg.stanford.edu/upload/V3LessonPlans/Executive%20Summary%2011.21.16.pdf
• Stieglitz, S. & Dang-Xuan, L. 2013. Emotions and information diffusion in social media—
sentiment of microblogs and sharing behavior. Journal of Management Information Systems,
29(4), 217–47.
• Thelwall, M., Buckley, K. & Paltoglou, G. 2011. Sentiment in Twitter events. Journal of the
American Society for Information Science and Technology, 62(2): 406–18.
• Vargo, C.J., Guo, L. & Amazeen, M.A. 2018. The agenda-setting power of fake news: A big
data analysis of the online media landscape from 2014 to 2016. New Media & Society, 20(5),
2028–2049.
• Van Duyn, E. & Collier, J. 2019. Priming and Fake News: The Effects of Elite Discourse on
Evaluations of News Media, Mass Communication and Society, 22:1, 29-48,
• Vosoughi, S., Roy, D. & Aral, S. 2018. The spread of true and false news online. Science,
359(6380):1146–1151.
Notas do Editor
Andy – how is emotional AI used in social media – sentiment analysis – Facebook link
Context to Cambridge Analytica: finding emotional triggers for each individual voter.
From PoV of those interested in moving image, interesting: trust in the kino eye, seemingly objective truths, documentary realism, trust…
The Joe Rogan Experience is a audio and video podcast hosted by American comedian, actor, sports commentator, martial artist and television host. Featured Elon Musk smooking a reefer.