Measures of Dispersion and Variability: Range, QD, AD and SD
Data Science: Case "Political Communication 1/2"
1. Introducing. . . Studying social media in poltical communication Conclusion
Fundamentals of Data Science: Case “Political
Communication”
Damian Trilling
d.c.trilling@uva.nl
@damian0604
www.damiantrilling.net
Afdeling Communicatiewetenschap
Universiteit van Amsterdam
12-09-2016
Fundamentals of Data Science: Case “Political Communication” Damian Trilling
2. Introducing. . . Studying social media in poltical communication Conclusion
Today
1 Introducing. . .
. . . the people
. . . the schedule
. . . the topic
2 Studying social media in poltical communication
Selective exposure and filter bubbles
Fragmentation
Polarization
Politicians on social media
Social media and public opinion
3 Conclusion
Fundamentals of Data Science: Case “Political Communication” Damian Trilling
3. Introducing. . . Studying social media in poltical communication Conclusion
. . . the people
Introducing. . .
. . . the people
Fundamentals of Data Science: Case “Political Communication” Damian Trilling
4. Introducing. . . Studying social media in poltical communication Conclusion
. . . the people
Introducing. . .
Stevan dr. Stevan Rudinac
Postdoctoral Researcher @ Intelligent Sensory
Information Systems // IvI // UvA
• received PhD degree in Computer Science @
TU Delft 2013
• graduated in Electrical Engineering @
University of Belgrade 2006
• worked @ NFI, TU Delft, TU Eindhoven and
University of Belgrade
• interested in multimedia information retrieval
with a focus on urban computing and security
applications.
s.rudinac@uva.nl Science Park 904 C3.253
https://staff.fnwi.uva.nl/s.rudinac/
Fundamentals of Data Science: Case “Political Communication” Damian Trilling
5. Introducing. . . Studying social media in poltical communication Conclusion
. . . the people
Introducing. . .
Damian
dr. Damian Trilling
Assistant Professor Political Communication &
Journalism
• studied Communication Science in Münster
and at the VU 2003–2009
• PhD candidate @ UvA 2009–2012
• interested in political communication and
journalism in a changing media environment
and in innovative (digital, large-scale,
computational) research methods
@damian0604 d.c.trilling@uva.nl
REC-C 8th
floor www.damiantrilling.net
Fundamentals of Data Science: Case “Political Communication” Damian Trilling
6. Introducing. . . Studying social media in poltical communication Conclusion
. . . the people
Introducing. . .
You
Your name?
Your background?
Your reason to follow this course?
Fundamentals of Data Science: Case “Political Communication” Damian Trilling
7. Introducing. . . Studying social media in poltical communication Conclusion
. . . the schedule
Introducing. . .
. . . the schedule
Fundamentals of Data Science: Case “Political Communication” Damian Trilling
8. Monday, 12 Sept., 11–13 (Damian)
Intro to our social science case:
Social media analysis in political communication
Tuesday, 13 Sept., 9-11 (Stevan)
The research pipeline
Working with the Twitter API
Data preprocessing and sentiment analysis
Tuesday, 13 Sept., 13-17 (you)
Project: It’s your turn!
Thursday, 15 Sept., 9-11 (Stevan & Damian)
Practical work: Helping you with the project
Thursday, 15 Sept., 11-17 (you)
Project: It’s your turn!
Thursday, 15 Sept., 17-19 (Stevan & Damian)
Presentations: Teams pitching the progress
9. Monday, 19 Sept., 11–13 (Damian)
Analyzing political content on social media: Examples of research so far
Tuesday, 20 Sept., 9-11 (Stevan)
Topic analysis, correlations, visualization
Guest presentation Joost Boonzajer Flaes, Twitter UK
Tuesday, 20 Sept., 13-17 (you)
Project: It’s your turn!
Thursday, 22 Sept., 9-11 (Stevan & Damian)
Practical work: Helping you with the project
Thursday, 22 Sept., 11-17 (you)
Project: It’s your turn!
Thursday, 22 Sept., 17-19 (Stevan & Damian)
Presentations: Teams pitching the final results
10. Introducing. . . Studying social media in poltical communication Conclusion
. . . the topic
Introducing. . .
. . . the topic
Fundamentals of Data Science: Case “Political Communication” Damian Trilling
11. Introducing. . . Studying social media in poltical communication Conclusion
. . . the topic
This case is about
social sciences
⇒communication science
⇒political communication
Fundamentals of Data Science: Case “Political Communication” Damian Trilling
12. Introducing. . . Studying social media in poltical communication Conclusion
. . . the topic
What is Communication Science?
• looks at communication between actors in society
• is one of the empirical social sciences
• (mainly) focuses on mediated communication rather than
interpersonal communication
Fundamentals of Data Science: Case “Political Communication” Damian Trilling
13. Introducing. . . Studying social media in poltical communication Conclusion
. . . the topic
Political communication
journalists citizens
political actors
Fundamentals of Data Science: Case “Political Communication” Damian Trilling
14. Introducing. . . Studying social media in poltical communication Conclusion
. . . the topic
Methods to study communication
qualitative
• discourse analysis
• interviews
• focus groups
Fundamentals of Data Science: Case “Political Communication” Damian Trilling
15. Introducing. . . Studying social media in poltical communication Conclusion
. . . the topic
Methods to study communication
qualitative
• discourse analysis
• interviews
• focus groups
quantitative
• survey
• experiment
• content analysis
• network analysis
Fundamentals of Data Science: Case “Political Communication” Damian Trilling
16. Introducing. . . Studying social media in poltical communication Conclusion
. . . the topic
Methods to study communication
quantitative
• survey
• experiment
• content analysis
• network analysis
Fundamentals of Data Science: Case “Political Communication” Damian Trilling
17. Introducing. . . Studying social media in poltical communication Conclusion
. . . the topic
The link with data science
• “Computational social science”
Lazer, D., Pentland, A., Adamic, L., Aral, S., Barabási, A.-L., Brewer, D., . . . van Alstyne, M. (2009).
Computational social science. Science, 323, 721–723. doi:10.1126/science.1167742
Kitchin, R. (2014). Big Data, new epistemologies and paradigm shifts. Big Data & Society, 1(1), 1–12.
doi:10.1177/2053951714528481
Fundamentals of Data Science: Case “Political Communication” Damian Trilling
18. Introducing. . . Studying social media in poltical communication Conclusion
. . . the topic
The link with data science
• “Computational social science”
• “In short, a computational social science is emerging that
leverages the capacity to collect and analyze data with an
unprecedented breadth and depth and scale.’ (Lazer et al.)
Lazer, D., Pentland, A., Adamic, L., Aral, S., Barabási, A.-L., Brewer, D., . . . van Alstyne, M. (2009).
Computational social science. Science, 323, 721–723. doi:10.1126/science.1167742
Kitchin, R. (2014). Big Data, new epistemologies and paradigm shifts. Big Data & Society, 1(1), 1–12.
doi:10.1177/2053951714528481
Fundamentals of Data Science: Case “Political Communication” Damian Trilling
19. Introducing. . . Studying social media in poltical communication Conclusion
. . . the topic
The link with data science
• “Computational social science”
• “In short, a computational social science is emerging that
leverages the capacity to collect and analyze data with an
unprecedented breadth and depth and scale.’ (Lazer et al.)
• “the computational social sciences employ the scientific
method, complementing descriptive statistics with inferential
statistics that seek to identify associations and causality. In
other words, they are underpinned by an epistemology wherein
the aim is to produce sophisticated statistical models that
explain, simulate and predict human life.” (Kitchin)
Lazer, D., Pentland, A., Adamic, L., Aral, S., Barabási, A.-L., Brewer, D., . . . van Alstyne, M. (2009).
Computational social science. Science, 323, 721–723. doi:10.1126/science.1167742
Kitchin, R. (2014). Big Data, new epistemologies and paradigm shifts. Big Data & Society, 1(1), 1–12.
doi:10.1177/2053951714528481
Fundamentals of Data Science: Case “Political Communication” Damian Trilling
20. Studying social media in poltical communication
Selective exposure and filter bubbles
21.
22. Introducing. . . Studying social media in poltical communication Conclusion
Selective exposure and filter bubbles
Avoiding dissonant information is human.
Festinger, 1956
• People tend to avoid cognitive dissonance
• One effective way: avoiding information that conflicts
pre-existing beliefs
Fundamentals of Data Science: Case “Political Communication” Damian Trilling
23. Introducing. . . Studying social media in poltical communication Conclusion
Selective exposure and filter bubbles
And it does happen in political communication.
Lazarsfeld, Berelson, & Gaudet, 1944
• Republicans are mainly exposed to the Republican campaign
• Democrats are mainly exposed to the Democratic campaign
Fundamentals of Data Science: Case “Political Communication” Damian Trilling
26. Introducing. . . Studying social media in poltical communication Conclusion
Fragmentation
Fragmentation
Sunstein, 2001 (and many others)
• People will only use those news media that cater to their
interest
• “echo chambers”
• Loss of a common core of issues
• Loss of democratic discourse
⇒ news avoidance, entertainment preference as predictor of news
use in the new media ecosystem (Prior, 2015)
Fundamentals of Data Science: Case “Political Communication” Damian Trilling
27. Introducing. . . Studying social media in poltical communication Conclusion
Polarization
Polarization
Selective exposure to ideologically congruent content
• If people don’t hear the other side any more, they become
more extreme
Fundamentals of Data Science: Case “Political Communication” Damian Trilling
28. Introducing. . . Studying social media in poltical communication Conclusion
Polarization
US Presidential Elections: Vote share 1952, 1956
http://xenocrypt.blogspot.de/2013/02/presidential-results-by-1952-districts.html
Fundamentals of Data Science: Case “Political Communication” Damian Trilling
29. Introducing. . . Studying social media in poltical communication Conclusion
Polarization
US Presidential Elections: Vote share 2008, 2012
http://xenocrypt.blogspot.de/2013/02/presidential-results-by-1952-districts.html
Fundamentals of Data Science: Case “Political Communication” Damian Trilling
30. Introducing. . . Studying social media in poltical communication Conclusion
Polarization
Let’s conclude. . .
• People are selective
• Nowadays, there is more media content to choose from
• content one politically agrees with ⇒ polarization
• entertainment over politics, only exposure to topics one is
interested in beforehand ⇒ fragmentation
Fundamentals of Data Science: Case “Political Communication” Damian Trilling
31. Introducing. . . Studying social media in poltical communication Conclusion
Polarization
The filter bubble
Pariser, 2011
• Algorithms increasingly guess what we might like and choose
for us (FB, Google,. . . )
• Even if we do not avoid actively, we are living in a filter bubble
Fundamentals of Data Science: Case “Political Communication” Damian Trilling
32. Conover, M. D., Gonçalves, B., Flammini, A., & Menczer, F. (2012). Partisan asymmetries in online political
activity. EPJ Data Science, 1(6), 1–19.
33. Introducing. . . Studying social media in poltical communication Conclusion
Polarization
How much of an echo chamber are social media?
Fundamentals of Data Science: Case “Political Communication” Damian Trilling
34. Introducing. . . Studying social media in poltical communication Conclusion
Polarization
How much of an echo chamber are social media?
“We estimated ideological preferences of 3.8 million Twitter users
and, using a data set of nearly 150 million tweets concerning 12
political and nonpolitical issues. [...] Overall, we conclude that
previous work may have overestimated the degree of ideological
segregation in social-media usage”
Barbera, P., Jost, J. T., Nagler, J., Tucker, J. A., & Bonneau, R. (2015). Tweeting from left to right: Is online
political communication more than an echo chamber? Psychological Science, 26(10), 1531–1542.
Fundamentals of Data Science: Case “Political Communication” Damian Trilling
35. Introducing. . . Studying social media in poltical communication Conclusion
Polarization
How much of an echo chamber are social media?
“We estimated ideological preferences of 3.8 million Twitter users
and, using a data set of nearly 150 million tweets concerning 12
political and nonpolitical issues. [...] Overall, we conclude that
previous work may have overestimated the degree of ideological
segregation in social-media usage”
Barbera, P., Jost, J. T., Nagler, J., Tucker, J. A., & Bonneau, R. (2015). Tweeting from left to right: Is online
political communication more than an echo chamber? Psychological Science, 26(10), 1531–1542.
“We find that users share news in similar ways regardless of outlet
or perceived ideology of outlet, and that as a user shares more
news content, they tend to quickly include outlets with opposing
viewpoints. [...] Specifically, users in our sample who sent multiple
tweets tended to increase the ideological diversity in news they
shared within two or three tweets”
Morgan, J. S., Shafiq, M. Z., & Lampe, C. (2013). Is news sharing on Twitter ideologically biased? Proceedings of
the 2013 conference on Computer supported cooperative work (pp. 887–897). ACM.
Fundamentals of Data Science: Case “Political Communication” Damian Trilling
36. Introducing. . . Studying social media in poltical communication Conclusion
Polarization
A first answer to the question why we should study social
media:
They might change the way people
are exposed to news and political
messages – which could lead to
fragmentation and polarization.
But we don’t have conclusive answers yet. . .
Fundamentals of Data Science: Case “Political Communication” Damian Trilling
38. Introducing. . . Studying social media in poltical communication Conclusion
Politicians on social media
The politician–citizen edge is now finally a viable way
. . . no need to take the detour through mass media any more.
journalists citizens
political actors
Fundamentals of Data Science: Case “Political Communication” Damian Trilling
39. Introducing. . . Studying social media in poltical communication Conclusion
Politicians on social media
Consequences
• politicians use social media to be more in control, bypassing
the journalistic filter
• reach other target groups
• but also: from one-way to two-way communication
Fundamentals of Data Science: Case “Political Communication” Damian Trilling
40. Introducing. . . Studying social media in poltical communication Conclusion
Politicians on social media
How politicians (should) communicate only
• interactivity
• personalization
can enhance political involvement
Kruikemeier, S., van Noort, G., Vliegenthart, R., & de Vreese, C. H. (2013). Getting closer: The effects of
personalized and interactive online political communication. European Journal of Communication, 28(1), 53–66.
Fundamentals of Data Science: Case “Political Communication” Damian Trilling
41. Introducing. . . Studying social media in poltical communication Conclusion
Politicians on social media
Effects of politicians on social media
• Using social media impacts voting, especially
“voorkeurstemmen”
Jacobs, K., & Spierings, N. (2014). . . . Maar win je er stemmen mee ? De impact van Twittergebruik door politici
bij de Nederlandse Tweede Kamerverkiezingen. Tijdschrift Voor Communicatiewetenschap, 42(1), 22–38.
Kruikemeier, S. (2014). How political candidates use Twitter and the impact on votes.Computers in Human
Behavior, 34, 131–139.
Fundamentals of Data Science: Case “Political Communication” Damian Trilling
44. Introducing. . . Studying social media in poltical communication Conclusion
Social media and public opinion
Things to keep in mind
• Be careful in generalizing!
• Often used because of easy to access API, but is it really the
right data source for your question? . . .
Fundamentals of Data Science: Case “Political Communication” Damian Trilling
45. Introducing. . . Studying social media in poltical communication Conclusion
Social media and public opinion
Example: Twitter-based prediction of election results
Fundamental flaws
• heavily skewed user base
• better than random does not mean better then more sensible
baseline (last election results, . . . )
• published after results were known
• arbitrary choices on what to include
• overly simplistic assumptions (e.g., number of mentions =
support)
Gayo-Avello, D. (2013). A Meta-Analysis of State-of-the-Art Electoral Prediction From Twitter Data. Social
Science Computer Review, 31(6), 649–679.
Jungherr, A., Jürgens, P., & Schoen, H. (2011). Why the Pirate Party Won the German Election of 2009 or The
Trouble With Predictions: A Response to Tumasjan, A., Sprenger, T. O., Sander, P. G., & Welpe, I. M.
“Predicting Elections With Twitter: What 140 Characters Reveal About Political Sentiment.” Social Science
Computer Review, 30(2), 229–234.
Fundamentals of Data Science: Case “Political Communication” Damian Trilling
47. Introducing. . . Studying social media in poltical communication Conclusion
Conclusion (1)
• The changing media environent can have impact on society at
large (fragmentation, polarization)
• It changes the communication triangle between politicians,
journalists, and the public
Fundamentals of Data Science: Case “Political Communication” Damian Trilling
48. Introducing. . . Studying social media in poltical communication Conclusion
Conclusion (2)
• Social media data can be linked to political outcomes
• But be careful to generalize!
Fundamentals of Data Science: Case “Political Communication” Damian Trilling
49. Introducing. . . Studying social media in poltical communication Conclusion
Questions?
d.c.trilling@uva.nl
@damian0604
www.damiantrilling.net
Fundamentals of Data Science: Case “Political Communication” Damian Trilling