04/20/2023: Duncan Watts

[GabeNicholson]

Comment below with a well-developed question or comment about the reading for this week's workshop.

If you would really like to ask your question in person, please place two exclamation points before your question to signal that you really want to ask it.

Please post your question by Tuesday of the coming week, at 11:59 PM. We will also ask you all to upvote questions that you think were particularly good. There may be prizes for top question-askers.

[Dededon]

Hi Professor Watts,
Thank you so much for offering this fantastic talk!
I'm particularly interested in Hosseinmardi and your paper of intake of radical content on YouTube. Social media research on YouTube is a brand new direction that are yet to be discovered by the scholarship, and I very appreciate the paper's idea of separating the on-and-off platform content consumption. Are there replicable methodolgies to record, measure, and evaluate the off-platform content consumption of any selected user group? That would be definitely helpful for future researches!

[JoeHelbing]

Hello Professor Watts,

Interesting the online "echo-chamber" phenomenon was less than the television, but theoretically makes sense given the smaller selection and greater ability to self select into those. Was there any efforts by tech companies creating the feeds to reduce echo-chamber effect potentially, aka is this a choice by those creating the algorithms to reduce that effect?

[borlasekn]

Hi Prof. Watts,

Thanks for sharing your expertise with us. I was wondering whether you think the same sort of "echo chamber" politics apply on platforms that are largely used for what people would consider "non-political" issues, such as on TikTok. In theory, people consume certain content and the algorithms build their explore pages based on the content they like. Now, assume an individual largely watches comedy videos. I believe that their are still echo chambers within these algorithms, and people will likely be swayed by the creators whose content they assume. For example, say an individual watches cooking videos and in particular certain people's content. If the content creator happened to be liberal, is that individual, by proxy, more likely to be liberal? How would you study these sorts of indirect and/or cascading echo chamber effects? Thanks!

[taizeyu]

Hi Prof. Watts,
Thanks for your speech. Have the trends identified in this study changed in recent years? Because we all know that the form of society will change with technological and social progress. Can this model be applied universally?
Thanks

[hsinkengling]

Dear Prof. Watts,

This is a fascinating paper! I wonder whether the difference in online versus tv news consumption could be attributed to the purpose people use these sources for (and whether we can control for that). For example, people might use online news more as a source of information (faster, more diverse) while tv is used more and more for political cheerleading (driven by political brand personalities). This would imply that not all consumption is equal: selective focus/concentration when consuming certain news sites might be more important than bland information gathering.

In addition to measuring polarization through consumption, how do these trends compare with data on public opinion, where polarization is more obvious? Are patterns of polarization driven mostly by hardline supporters?

[awaidyasin]

Thanks for sharing your work, Professor! You show that biased news consumption is not persistent, i.e., people consuming left- or right-winged news are less likely to consume that in the following period. I guess what's also important is not just the political inclination of news but also who is delivering it. For instance, some people might simply like Tucker Carlson and want to hear him deliver news. Does your empirical evaluation include some kind of person fixed-effects?

[XTang685]

Hi Prof. Watts,
You analyze the news consumption of web and TV audiences to understand how partisanship influences media choices. The study finds that while web audiences consume more partisan news than TV audiences, both groups tend to favor news sources that align with their political beliefs. The study also reveals that the online media landscape is more fragmented than the TV landscape, with fewer dominant news sources. My question is, how can we address the issue of algorithmic bias in online news consumption, considering the fact that the personalization of news feeds often reinforces users' existing political beliefs and contributes to the formation of filter bubbles, leading to a lack of exposure to diverse viewpoints and potential polarization?

[JerryCG]

Dear Prof. Watts,

One thing I am concerned with is it seems hard to purely classify a person as TV viewer or online viewer. I think it might be for most people they are hybrid viewer of both TV and online channels. Is it possible that someone is partisan-segregated on TV but not so online? or the other way round? And I think the TV program resources are much limited than online resources. Thus could this phenomenon just be driven by the diversity of resources so online will have less segregation?

Best,
Jerry Cheng

[Hongkai040]

Hi Prof. Watts, thank you for sharing your work! My question is what is the future of democracy in US in a digital age, like a decade or two decades from now? It's looks like a big question, but the rationale behind the question is based the findings of your papers: Your papers argues that TV, a shrinking market, contribute the most political segregation, while the situations on social platforms are more moderate. We may safely assume that the younger generations are shifting from medias like TV to Internet streaming platforms and social medias. If this is the case, does that mean that the society is not becoming more polarized in the future when the generation watching TV fade away from the stage?

[ShiyangLai]

Dear Prof. Watts,

Thank you for sharing your research on TV-driven partisan audience segregation. Your descriptive analysis is very interesting, and it is surprising to see such a fast drop in partisan segregation for both TV and Web news audiences. While I understand that your data is not suitable for identifying the drivers of desegregation, I am curious about your hypothesis regarding this phenomenon.

I am also a bit confused about one thing: how do you define TV news audience and Web news audience? It's common to see people consuming news in a mixed way, so can we really say that an individual is partisan-segregated based solely on their content selection behavior on one platform? For example, what if someone chooses to watch specific types of news on TV but looks at very different content online? This situation is not uncommon. Personally, I sometimes watch TV with my family and choose something that they are interested in, but when I'm online alone, I may look at something that only interests me. I believe many young people do the same thing. So, can we really draw a conclusion about someone's level of partisan segregation based solely on their content selection behavior on one platform?

[pranathiiyer]

Thank you so much Professor Watts for sharing your work with us! How do you anticipate that this will change with age as younger generations get older? Given that your population varies with age, will we see trends more within the online space as opposed to television as younger generations get older and newer generations consume more and more news online?

[zoeyjiao1104]

Professor Watts,

Thank you so much for sharing your work.

Considering the growing trend of partisan news consumption and its potential impact on political attitudes and behaviors, what measures can be taken to mitigate the potential negative effects of such news consumption on individuals and society as a whole?

[erweinstein]

Hi Professor Watts,

Thanks for giving this talk to our workshop!

A key point you and you co-authors make is that "...to the extent that Americans do consume news, it is overwhelmingly from television, which accounts for roughly five times as much as news consumption as online". The relevance of this fact to much research on media consumption patterns, including the papers you note which overreact in their attempt to "debunk" the concerns about echo chambers, seems so obvious, yet it was not quite as obvious as it seemed, since it was being ignored. Overall, your point about the relative impact and importance of TV compared to online news reminds me of Hortaçsu and Syverson's 2015 work showing that for all of the emphasis on e-commerce, warehouse clubs like Sam's Club and Costco were at least 50% more impactful in terms of how much consumers buy from each type of store (at least in the 2010s; furthermore, their work also parallels yours in that a person is extremely likely to use both of, say Amazon and Costco, but not necessarily or likely in equal proportion, and that proportion matters). Or the fact that less than 40% of Americans have a 4-year college degree (although this number is higher than it used to be), while so much of "the discourse" (hyper-focused as it is on free speech/platforming questions, campus politics, and related matters) presupposes that attending a 4-year college is a universal shared experience for all who are reading/listening/watching. I hope that, like the work of Gelman et. al. that corrected some very-widespread (and commonly repeated by the media) misunderstandings about American partisan voting behavior, the work that you and your colleagues have done here rapidly becomes well-known and the new baseline for work on this topic.

[Yuxin-Ji]

Dear Prof. Watts,

Thank you so much for sharing your work with us! I really appreciate your research comparing the partisans in TV and web news. While the results show that there is a larger partisan in TV news audience, there might be selection bias among different audience groups of TV and web news. I wonder whether you would conclude a causal (or reverse causal) relationship between partisan and TV news?