China's Great Big Holey 'Twitter' Net
It is supposed to be the social media-powered enabler of free speech in China. However, Sina Weibo's reputation is under a microscope following a new study showing more than half of user accounts surveyed were empty, and there weren't many postings of original content. In a country with so much repression, though, maybe a few voices being heard is better than silence.
03/23/13 5:00 AM PT
A recent study by researchers at the University of Hong Kong suggests that China's Twitter-like Sina Weibo -- which boasted in February that it had surpassed 500 million users -- might be overrun by zombie accounts.
The study, which looked at a random sample of 30,000 Sina Weibo accounts, revealed that 57 percent of users had nothing, nada, in their timelines. The study also showed that just 0.5 percent of users posted more than 20 messages in a given week.
These findings don't seem to support the idea, espoused often, that Sina Weibo is a game changer for free speech in China.
In this TechNewsWorld podcast, we talk with one of the authors of the study, King-wa Fu. Fu talks about why previous studies of Sina Weibo overlooked the tumbleweeds blowing across so many accounts, and how his study fits (or doesn't fit) the narrative that Sina Weibo is a revolutionary tool for muzzled Chinese citizens.
Download the podcast (18:51) or use the player:
Here are some excerpts from the podcast:
TechNewsWorld: Let me ask you what was the impetus, or the motivation, for conducting this study? Were you skeptical of the numbers you were hearing about Chinese social media use? What was it that motivated you?
Fu: I think it is purely curiosity. We know there is a major focus and global attention on new media in China. So we tried to get more empirical data about how Chinese people are using this service.
We found out that we couldn't find very good, reliable, representative statistics on (Sina Weibo) use in China. For example, we tried to find out some usage statistics -- some of them are collected by using surveys. Normal survey data is skewed; a small sample is not very reliable. Some of the people tried to understand by using some data mining processes, to look at the data and try to analyze data. But we found out that most of them are basically using what I call non-randomized sampling approach. That means they are more focused on those people who are more likely to publish their Weibo. So, more likely to publish their Weibo are more likely to get sampled.
We found that there's maybe a bit of bias in these kinds of sampling approaches. There's a lack of reliable data, so that's why we wanted to collect (this information). So we deployed what's called the random sampling approach. Just imagine every Weibo user carries a 10-digit number -- just like a telephone number. We used this unique number to deploy the randomized approach. We tried to make a "dial" to all these people by using a computational method. That dial means we just collect data from their timeline.
So we collect data and then use analysis to try to make sense of their pattern of usage and how many percent of people contribute to how many percent of the data. And also try to find out the predictor for the people who have a high number of posts. So we had a lot of questions that we wanted to find out about this microblog.
TechNewsWorld: Now, in addition to finding that 57 percent of the accounts you followed had nothing in the timeline, you also found that about 87 percent of the users had no original posts, and that about 89 percent did not repost any original content from another account. So, your findings make it sound like there's maybe not as much activity on Sina Weibo as people thought. Were you surprised by the lack of activity that you saw when you started studying these 30,000 accounts?
Fu: A bit surprised. We expected to have a high percentage of people -- basically they are very passive. They are basically just reposting other people's content as opposed to writing their own content. But we didn't have any idea about what percent of people have this characteristic. We understand that a lot of people just read other people's posts, just repost other people's posts. But we didn't expect that to go up to almost (nine out of 10) microblog users in China.
TechNewsWorld: The Wall Street Journal, for instance, wrote, "The study reveals Weibo to be less like a town square and more like a speaker's corner in Hyde Park, where a vocal minority dominate the discussion." And this is an interesting idea, especially because Sina Weibo has been heralded as a revolution for China.
In August 2012, for instance, there was a Yale research paper called, "How Weibo is Changing China," and then a few months later NPR had a radio piece called, "How Ordinary Chinese are Fighting and Talking Back," and this was all about Weibo and social media platforms. Do you think that your research kind of casts doubt on the revolutionary power of Sina Weibo? Do you think people were maybe too excited about the potential for this compared to what it's actually doing?
Fu: I don't think it's really a revolutionary tool. But it's a very important tool for Chinese people to express their opinions in the public space. I don't really agree 100 percent with (theJournal's) report on using that public park example to look at Sina Weibo. Because if you just look at -- yeah, you're right, it's just a small portion of people in China who are able to have a presence on Sina Weibo. But you must understand in Chinese society, information is mostly censored. That means that if this small group of people don't have Weibo, they don't make use of the Internet, then they have no way to express their opinions in other channels. For example, they have no right to form public assembly. They can't organize a meeting in the park. They can't write a letter to the editor of The People's Daily. They can't appear on television.
So in China, Weibo is a very unique and important tool for them to express their opinion. Even though there's a very small number -- it's still very influential.