02/04: Yinxian Zhang

[smiklin]

Comment below with questions or thoughts about the reading for this week's workshop.

Please make your comments by Wednesday 11:59 PM, and upvote at least five of your peers' comments on Thursday prior to the workshop. You need to use 'thumbs-up' for your reactions to count towards 'top comments,' but you can use other emojis on top of the thumbs up.

[wanxii]

Thank you so much for engaging us in this really intriguing topic and I believe it certainly deserves illuminating debates and further researches. For the study itself, I might question if it's sufficient to attach such prominent importance to those Weibo "opinion leaders" defined in the papar and state that "If this group does not champion democracy and liberal values, then society at large is either unlikely to.":

1. One reason is that social meida platforms, like Weibo have long been accused of taking advantages of fake followers, comments and likes to boost accounts' popularity. If we use certain kinds of fake-follower detectors or inspect comments under the posts, we might discover a dramatic shrinkage in the follower numbers or identify a large portion of content-irrelated replies. Therefore, I wonder if you have taken more comprehensive methods to validate the social influence of those accounts.
2. Speaking of social influence, we might admit that people like to hear what they want to hear. Therefore, those opinion leaders might mainly attract followers sharing exactly the same political ideology. This kind of users polarization could potentially lower their influence on a wider range of population.
3. Another issue is concerning different users' demographic features accross social media plafroms. You've also mentioned in the paper that Weibo user samples certainly diverge from the total Chinese population. Compared to Weibo, platforms in China like Zhihu are recognized as gathering places for public intellectuals (so-called Gongzhi). Therefore, it might be insufficient to conclude so far whether various internet communities share similar opinion shift patterns and determine which platforms dominate the mainstream and exert more influence on the public. I wonder how you'd address those problems to support your conclusions. Thanks!

[jsoll1]

Thanks for sharing your working paper with us! I'm looking forwards to the presentation!

I have to wonder, why did you decide to select a set of opinion leaders to follow for the entire time period? Might it not have been better to break it up, as different people may have been more influential in different years? Unlike you I don't have access to the full data, but I can imagine a particular opinion leader falling out of the public eye due to having an unpopular opinion in a particular year.

[yierrr]

Thanks for the enlightening research! Similar to @wanxii , I also have concerns regarding the data and the sample used in the paper: 1) the words on which the tf-idf analyses were performed all seemed to be quite formal and standardized, but are also very likely to be the ones censored by the system. Hence, in order to post something, I've seen many people replacing these sensitive words with wrong characters of the same pronunciations, adding random punctuation marks in between, or even posting screenshots of what they wrote. Hence, by searching these standardized words, what's left is highly likely to be more flattering than critical (especially with the censorship tightening). How did you address this concern? 2) regarding the sample of opinion leaders, I noticed that almost 90% of them are male and more than half of them reside in Beijing, and they would be presumably well-off - this may indicate the majority of them can be part of the most privileged by the system, which certainly would make it more difficult for them to speak for systems that might make them less privileged (and in some times more than others). In addition, it seems that the sample was generated by filtering opinion leaders keen on politics, but people who can be famous on political views on Weibo can be biased in the first place - for example, females are never trusted with political opinions and dissidents could have their accounts deleted instantly. Therefore, what do you think of the bias this sample may induce, and whether it might be more valuable to study, for example, foodie accounts beginning to make comments about the politics as better reflections of "voices of the people"? Thank you!

[bakerwho]

Thank you for this fascinating paper with a perspective that isn't commonly seen!

I'm curious about how these opinion leaders fit into the information ecosystem. What are the popular sources of information in China for topics like world politics (such as the US or UK elections)? Do opinion leaders also serve as educators and informers, or are they primarily concerned with discourse? Do they tend to summarize and characterize issues? What sources do they draw from? Are the Chinese media nationally owned, or are there private players? Are sources like Reuters, Associated Press, or perhaps Al Jazeera or the Guardian, popular?

I would expect there to be a divide between nationalists and liberals along these lines, with liberals trusting foreign sources more. Given China's well-documented suppression of internet freedoms, isn't it possible that conflicting viewpoints are simply less accessible for public discourse? How do we account for this effect in our discussion of passive legitimacy? Is it possible that this is a subverted form of populist legitimacy (where contrarian voices are structurally designed to be less heard and seen)?

[Rui-echo-Pan]

Thank you for sharing such exciting research with us! I strongly agree with you that "changes in public opinion during the pandemic mirror long-term trends in public opinion over the last decade"! At the first glance at the abstract, it comes to my mind that your finding could be also interestingly applied in the COVID-19 period during which the significant events in China and other countries, either positive or negative, influenced people's opinion on the democratic system a lot. And not surprisingly, you emphasize the focus on COVID-19 at the end of the paper!

But analysis on the public opinion change during COVID-19 is quite brief and general, so I'm thinking if further research could be developed in some specific ideas (possibly as another research): (1) How the public opinion would react to the significant events in both China and other countries, including both the positive and negative events. This might be similar to Shan & Chen (2020), but instead of exploring the trend, the response to significant events may be more focused on. (2) In your paper, people's public opinion is affected by the "news" on the outbreak in other countries. I think the narrative and interpretation of the news by the mainstream news provider, rather than only the fact itself, is also important to influence people's opinion. Because I suppose most people (especially most users on Weibo) form their views based on interpreted facts. It would be interesting to see how the opinion leaders'/ people's posts are related to the official narrative about the effective response at home and the outbreaks in other countries.

Looking forward to seeing the final draft!

[j2401]

Thank you for sharing with us. I have similar concern as @wanxii. Besides the issue of fake followers and comments, I would like to add another minor point: Have you considered other possible incentive for elite users to publish opinions related to politics or similar topics(which are regarded as sensitive topics) ? Weibo is actually subsidizing opinion leaders by awarding "interactions"(likes, comments, reposts), especially those with large volume of followers. The size of the subsidy is around 1-2 billion as of 2018. This is very likely to encourage opinion leaders posting conflicting materials that attracts more attention of public, for the purpose of money making, instead of real opinions exchange . I wonder how does this affect your sampling choice?

[chrismaurice0]

Thank you for sharing your work with us! My question is a simple one: for a whole paper on modern opinions of democracy in China, why is there only one mention of the Tiananmen Square protests tucked into the following sentence: "The desire for democracy and discontent with the government were translated into mass protests, as seen in the 1978 Democratic Wall Movement and the 1989 Students’ Movement."

Additionally, I think a great addition to this paper would be to look at social media posts from opinion leaders (or as many posts as you can collect) around the June 4th anniversary of the protest and analyze any patterns you see in the data.

[Bin-ary-Li]

Excellent papers! While it might be the case that the Chinese people on the internet are becoming more and more disillusioned with classical liberal democracy. Do you think that the state might play a role in deliberately promoting/fostering such an environment that will facilitate this process?

[JadeBenson]

Thank you so much for this fascinating research!

I would be interested in hearing more about polarization in opinion leaders posts. You mention that polarization between positive and negative terms has become more pronounced since 2014. But overall, it seems like there's been a confluence of opinions where "both sides" are growing more skeptical of democracy - increasing passive legitimacy. US politics and especially online public discussion seem to be characterized by polarization - what would you say have been the dynamics of polarization amongst Chinese opinion leaders? Perhaps related to this, have you noticed the rise of conspiracy theories or opinion leaders that focus on this type of content?

[yiq029]

Thank you for sharing your interesting research! Since Weibo has a quite strict censorship. How do you think about the value of studying the filtered posts from Weibo? Thank you.

[wanitchayap]

Since you have a good number of opinion leaders, do you see any potential future analysis using computational methods beyond content analysis at the aggregated level? For example, maybe a combination of content analysis with network analysis (each opinion leader as a node maybe?). Do you have any plan for this? Which method (or a combination of ones) one which question to explore do you think would be a promising one?

[ChivLiu]

Thank you for sharing this presentation! I think it is very interesting to use social media data from China since a lot of people are questioning the accuracy of the data. I think people's opinion is strongly correlated with the information they receive. If they receive more positive information, they would also provide more positive feedbacks. If they only see complaints and wealth gaps, they would feel anxious and express the extremist inside. I wonder would you think that if a Weibo user follows a lot of users, who usually post positive terms, he or she would also post some positive opinions?

[Yutong0828]

This is a very engaging study! I was wondering if you investigated the background of the opinion leaders and their followers? As presented in the table 2, the opinion leaders are mostly males from Beijing, and their followers may also have an inclination to be more passionate about such topics, while the most of the netizens on Weibo may not really follow these opinion leaders. So, how would you prove that analysis of democracy-related opinion leaders on Weibo can reflect general trend of people's attitudes toward democracy? Thank you very much!

[bowen-w-zheng]

Thanks for the paper! My questions are similar to Wanxi's. I wonder if opinion leaders on Weibo can be thought of as a good thermometer for the general public's opinions. Weibo has a history of cooperating with government agencies to censor opinions against the regime and opinion leaders that are pro-democracy might shift to other platforms like Twitter. Looking forward to the presentation and how you address these concerns.

[boyafu]

Thank you for sharing this fascinating research! I was wondering if the conclusions would keep robust among different online social platforms in China, since the censorship and mainstream topics may differ among different platforms especially in an online environment. Thanks!

[alevi98]

Thank you Professor Zhang for sharing your research with us and speaking at our workshop! Indeed, people in the U.S. really like to talk about the decline of democracy worldwide, so this is very thought-provoking and timely.

I'd be curious to hear your thoughts on what Margaret Roberts calls "porous censorship," wherein the Chinese government specifically targets gatekeepers of information such as journalists, public academics, prominent figures etc. who are prone to criticizing the government's censorship practices or inciting collective action, but uses more subtle mechanisms to spread the slow of inconvenient information among the general population? Along the lines of others, I'm wondering how government censorship has impacted the results of your analysis.

I think the idea of a "porous censorship" strategy that targets folks in positions similar to what you call "opinion leaders" would have profound implications if it systematically removes gatekeepers of information who are critical of an authoritarian regime.

Of course, as someone born and raised in a Western country, that's my intuition because of my bias. People in the U.S. are having a lot of conversations internally about what restoring democracy and returning to our first principles truly means in an era like today. Some people don't think it's possible, but others want to give the system another chance and right many of the wrongs that cause people to criticize democracy, such as reparations policies that would offset legacies of colonialism, slavery, and exploitation. Is it the system that's broken or what we have internalized as "normal" in the system?

I also think it's interesting that the most radical factions in the U.S. fracture from democracy on the basis of freedom. Anarchists on the far left believe a social contract deprives them of the freedom of their natural state, and in a strange way, many people on the alt-right believe the conditions of a globalized, multicultural world are destroying the basic liberties of what they consider "their" nation. Do you believe cultural relativism is at play here; that there is no one "right" system but simply political values instilled across generations?

[rkcatipon]

Thank you for sharing your work with us, Prof Zhang! It's great to see content analysis methods used on non-English language.

Many of my questions have been echoed (and better put 😄 ) by others. I am curious if the disillusionment of these thought-leaders may also be in response to network engagement? Is it driven by changes in the community or the thought-leader groups?

[YileC928]

Hi professor Zhang, thank you for presenting your paper! I have a small concern regarding the implication of the opinion leaders' behavior to the whole population. Why do you think "their attitude toward democracy has significant implications for the future of China" and "If this group does not champion democracy and liberal values, then society at large is either unlikely to" (page8)?

[YijingZhang-98]

Hi Professor, thanks for your work. I really enjoy reading it. I also observed the phenomenon that the "Chinese people’s rosy expectations of model democracy were challenged by increasing exposure to perceived subpar performance and racial discrimination of democratic regimes. This, in turn, strengthens popular support for the Chinese government". And thank you frame and theorize it as "passive political legitimacy of the Chinese regime."

I found an announcement from Guangdian in Jan. 31, which says the unauthorized account could not post news or discussion about the current polical affairs. You may find it here https://www.sohu.com/a/447903401_613537?spm=smpc.author.fd-d.2.1612101600032Zw5NOp7. It might be interesting if you could discuss the new regularization.

[YuxinNg]

Hi Professor Zhang! Thanks for sharing your paper. Like many others, I am also concerned about the censorship problem. Many platforms in China have been censored and the strength of censorship across different platforms are different. Do you think this will be a problem for your research? How are you gonna deal with this. Thanks!

[nwrim]

Thanks for a fascinating paper. Similar to Regina's words above, all of the questions that I had were asked by my peers above, so I will just ask a very vague question to get more context: How can abridge the gap between social media's opinions (Weibo in this paper) and the actual society? As I am very, very ignorant of the social media landscape of China, I wonder the results in the cyber world can directly be a proxy of what is actually happening in the real world. As people above mentioned, there are problems like censorship, the loud minority, population biased to favor certain groups, etc.

[egemenpamukcu]

Hello Professor, thank you for the insightful paper.

Chinese people’s rosy expectations of model democracy were challenged by increasing exposure to perceived subpar performance and racial discrimination of democratic regimes.

I am from Turkey and I feel like this statement does a really good job of describing the meainstream sentiment towards the 'West' by the pro-government portion of the population. I also feel like there is an unfair exaggeration when it comes to the 'shortcomings' of western democracy in authoritarian regimes or less mature democracies with weaker democratic institutions. Do you think this is the case, and if you do, do you think the way these shortcomings are portrayed ends up hurting the effort towards democracy in these countries, or, from a more optimistic point of view, can it inform the effort towards democracy and make it strive for a higher standard of justice, fairness, and equity?

[xzmerry]

Thanks for sharing this paper! After reading your paper, I feel that my understanding of 'democracy' is close to the 'new democracy' mentioned by Mao (which is the rule of people). Actually, as a Chinese, I sometimes cannot understand why western countries say that we are not democratic enough, and I am unclear what types of democracy we should fight for (or why it is meaningful to do so). So I feel that most Chinese people may have an inaccurate understanding of democracy or at least different from the definition prevalent in most western countries (as you have mentioned in the paper).

So my question is related to the measure: how you ensure that most opinion leaders on Weibo have a correct understanding of democracy as well as their readers?

As some of your analysis is based on posts that contain the keyword “democracy”, so I just wonder whether this could accurately reflect people's true attitude towards democracy. Because certain changing patterns in your research might be caused by people's changing understanding towards 'democracy', the term itself.

Besides, although opinion leaders are powerful shapers of public opinion compared to average netizens, their readers might be segmentalized into several relatively independent groups. So besides the total number of posts, how the opinions are diffused among readers is also important. The possibility that some so-called opinion leaders are actually net celebrities that write things to cater to the taste of more segmentalized readers is not be eliminated yet.

Again, thx for sharing!

[sabinahartnett]

Thank you for sharing your work! Similar to many classmates - I'm hoping you could elaborate a bit more about the role of censorship in these types of studies and additionally, how you think this could inform studies of more strictly monitored content (esp. given the recent decision by multiple platforms to remove accounts more liberally)

[NaiyuJ]

Hi Professor Zhang,

Thanks for sharing your excellent work! I really appreciate your combination of computational methods and qualitative analysis in studying such a public opinion topic.

I have a question from the perspective of politics about the role official propaganda plays in changing people's minds about democracy. In this article, the outcome you observe is that popular attitudes towards democracy are changing from the positive side to the negative side. The reason you suggest is that citizens get more exposure to the information and news about democratic countries through unofficial channels. That is very true from my own experience. But it seems that official propaganda also plays an important role in changing people's minds about western countries and democracy. Compared to the last few years, the state now continues to "advertise" the official ideology that authoritarianism is the most appropriate form in China and intentionally facilitate people's national consciousness towards the country and the sense of belonging to the community. For example, recent speeches of the Ministry of Foreign Affairs may directly critique the west.

And also, it is very important to find out the intentions that opinion leaders post their posts. If they post ideas that they truly believe, then the study is valid. But if they post ideas for financial purposes or to cater to government demands for potential benefits, how do we interpret the outcome?

[MkramerPsych]

Dr. Zhang,

Thank you for sharing your research with us! I, along with many of my classmates, am curious about potential confounds in Chinese social media data due to the active censor role played by the government. On a more technical level, I'm particularly curious if there is an established methodology for vetting social media data that may be influenced by censorship. Could there be a method for calculating the extent to which these influences bias the data?

[siruizhou]

Thank you for sharing this amazing work. I think Weibo is getting more and more censored and profit-driven over time. The recommendation algorithm might largely influence the popular contents. How do you think about the changing algorithmic impact?

[Yilun0221]

Thank you for the presentation. My concern is about the population of online discussions. For example, data shows that most people using Weibo, a popular social media platform in China, do not have a bachelor's degree from universities. I wonder how this kind of bias is explained in the results of this research?

[MengChenC]

Thank you for sharing your work, it is an interesting approach to explain the evolution of Chinese people's feeling to democracy. I am curious about how you justify or mitigate the effect that some posts on Weibo might have been censored and removed due to their touchy content, which may not offer a comprehensive inference of the opinion trend. Thank you.

[linghui-wu]

Thank you for presenting your work! I wonder how you could incorporate the dynamics of opinion leaders in the Chinese internet. To be specific, I am extremely interested in how the liberal leftist began showing their support for Trump in the latest years and in the 2020 election.

[NikkiTing]

Thank you for sharing your work! Your analysis is very interesting! It appears that the concept of democracy is too closely associated with the West, hence, negative impressions of the West have translated into negative sentiments of democracy. However, I wonder if this can really be understood as the public being against democratic reforms (e.g., in terms of freedom of speech). I would like to ask whether you think there would be a different outcome if the study focused on public opinions of reforms instead.

[skanthan95]

Thank you for sharing this work! One question I have is, why you chose LDA over NMF for generating topic models. Were the Weibo posts longer than the average Twitter post, since from what I understand, NMF tends to be used for documents with fewer words?

[hhx2207061197]

Hi, Professor Zhang.

Thanks for your presentation. Just want to know how do you think about the value of such research?

[a-bosko]

Thank you for sharing your work!

It is very interesting to learn about the changes in popular attitudes towards democracy in China. It is also interesting to see that there are increasing doubts and disillusionments with democracy in online discussions. As mentioned in the paper, democracy can be seen as a vague term, and there can be different subsections of democracy that are vastly different from one another.

Do you believe that this negative view of democracy will change in the future? As mentioned, the number of Chinese students studying in the U.S. has nearly quadrupled since 2010. This might help shift the viewpoint from negative to neutral, or even positive.

[ydeng117]

Thank you for sharing your work! the year 2013 is a very special one for China. President Xi promoted the ethos and idea of the "Chinese dream" and initiated another anti-poverty program. How may these strategic decisions affect your results?

[Lynx-jr]

Thank you for sharing this fascinating research! I have a question same as others, that is if the conclusions would keep robust among different online social platforms in China? Thanks!

[bjcliang-uchi]

Interesting topic. Not sure whether you pay attention to the current discussion on Chinese "pro-Trump" public intellectuals. See Lin Yao's "Beaconism and the Trumpian Metamorphosis of Chinese Liberal Intellectuals":

Political beaconism grows from the traumatizing lived experience of Maoist totalitarianism, sanitizes the West and particularly the United States as politically near-perfect, and gives rise to both a neoliberal affinity and a latent hostility toward baizuo.

That is, the definition of democracy in the Chinese context, even for liberal intellectuals, could be very skewed and diverged. Also, topic modeling approaches assume great stability in word meanings across contexts and time, and this assumption has been seriously violated as China has been trying to redefine human rights and democracy especially since Xi ("Tell a nice Chinese Story"). I am curious about how you address this language shift.

[LFShan]

Thanks for sharing!

In the past years, I have always felt that many Chinese' attitude towards a democratic government has shifted from positive to skeptical. It is enlightening to see that this phenomenon is being captured using the computational method. However, because many online communities' population is skewed towards younger generations, do you think the whole picture might differ from the online communities?

[Tanzi11]

Thanks for the riveting paper, professor Zhang! Like my classmates, I also wanted to enquire if these perceptions and attitudes change depending on social platforms and if governments actually affect public sentiment in terms of shaping their mindset?

[hihowme]

Thanks a lot for sharing this interesting paper and research representation! My question is that does the result and conclusion also applies to other social platforms? Thanks a lot!

[yutianlai]

Thanks for sharing! Could you please explain more on how you deal with possible selection bias in your results?

[Leahjl]

Thank you for sharing this interesting research! I'm curious how do you test the model on different social platforms?

[lulululugagaga]

Thanks for sharing your paper! I also observed that global events, especially events on American politics, often lead to heated discussion on hypocrisy and democracy. Why is the case?

[afchao]

Thank you for sharing your work with our group! I'm curious about the basis of the claim that political systems are reduced, in the public understanding, to a simple dichotomy between liberal democracy and authoritarianism. Is this specific to China? Or a broader statement about how people interpret ideology?

[FranciscoRMendes]

I am curious about sources of liberal information in China. It is seems that contrarian voices are suppressed by design, therefore the conclusions drawn from your paper are going to make this an interesting discussion. What do you think about applying this social media in a true democracy, like India?

[vinsonyz]

Thanks for sharing your paper with us, Yinxian! I am curious about that whether it is also true for comments on the Apps such as Toutiao.com where most of the contents are not uploaded by users.

[JuneZzj]

Thank you for presenting. I am interested in the time periods you chose from 2009 to 2017. Do you have any preference on the time period selection? Thank you.

[Jasmine97Huang]

Thank you Prof Zhang for this very interesting paper on a topic that many of us can deeply relate to. I notice that your data were all collected on Weibo. Is excluding western social media a conscious choice? I think a large amount of people under the Chinese diaspora opt to express their political opinions via western platforms, especially when they involves anti government sentiments. I also think that because the perception of western democracy is influenced by the information and news people receive, it may be interesting to analyze major Chinese news channels to understand what types of international news are people being exposed to. Major Chinese news media which are heavily controlled by the state would happily omit positive coverage and headline negative ones to evoke a downward comparison with democratic regimes. Lastly I'm interested to hear your opinions on the role people who have long-time and directly exposure to western ideologies play in shaping the perception of democracy at home?

[Dxu1]

Thank you for a fascinating paper! I am interested in the data collection and their potential implication. How would you evaluate the impact of the posts that are censored (not published, not censored after publish), particularly around sensitive leaders (current not the past leaders), and around sensitive years (2013)? Thank you very much!

[YaoYao121]

Thank you for sharing your work! I think just as some other students' question, data from social media like Weibo in China is really not as persuasive as, for example, Twitter. Could I ask what do you think about the representativeness of the data, please?