彩民宝典

The majority of Twitter users are unaware that researchers freely collect and analyze their tweets鈥攊ncluding deleted ones鈥攊n the name of science.听

Most believe that this shouldn鈥檛 be allowed without their consent. And many wrongly assume it would be a violation of Twitter鈥檚 terms of service, according to a new study by researchers from 彩民宝典 and University of Kentucky.听

鈥淭here is a ton of research right now using Twitter and other social media data,鈥� said lead author Casey Fiesler, an assistant professor in the Department of Information Science at 彩民宝典. 鈥淵et our study found that the majority of users may not even be aware that this is a thing that happens.鈥�

The comes at a time when several high-profile controversies鈥搃ncluding this week鈥檚 revelation that Cambridge Analytica used Facebook data to develop political ads鈥揳re raising questions about what鈥檚 ethical and legal in the burgeoning world of big data research.听

It is the first in a series of studies Fiesler is working on related to the ethics of such research.

For the study, Fiesler and co-author Nicholas Proferes, an assistant professor at University of Kentucky, surveyed 268 Twitter users, average age 32. The average participant had posted about 2,000 tweets and followed about 350 accounts.

When asked if they knew researchers sometimes used their tweets, 62 percent said 鈥榥o.鈥� When asked whether they thought researchers were permitted to do so without asking them, 43 percent said 鈥榥o.鈥� Of those, 61 percent said they believed the researchers would be breaking ethical rules, and 23 percent said they believed Twitter鈥檚 terms of service forbid it.

In reality, Twitter states in its privacy policy that it 鈥渂roadly and instantly disseminates your public information to a wide range of users, including search engines, developers, universities, public health agencies, and market research firms鈥︹��

When asked how they felt in general about Tweets being used in research, just 20 percent said they were generally uncomfortable with it and only 30 percent said they would opt-out of having any of their tweets used in research if they could.

More than half said that if a researcher contacted them to get consent and told them what the study was about, they would probably give it.

鈥淭hey were more upset about not being told than about the fact that it was happening,鈥� said Fiesler. 鈥淎nd they cared about things like what the study is about.鈥�

Levels of discomfort varied broadly depending on what information was used and how.听

For instance, 63 percent said they would not want deleted comments used for a study (as they often are), and 55 percent said they would not want one of their tweets, with their username attached, quoted in a published research paper. (In at least once instance, a reporter contacted the owners of such tweets to interview them). In contrast, if a tweet was analyzed along with millions or quoted anonymously, most users didn鈥檛 object.

Fiesler and Proferes suggest researchers consider such factors when designing their research.听

鈥淚 certainly don鈥檛 want to suggest we shouldn鈥檛 be doing research using Twitter data,鈥� said Fiesler. 鈥淏ut I do think that we should be thinking about how to do it ethically.鈥澨�

She pointed to several cases in which big data research has sparked public outcry:听This week, numerous news outlets reported that data gathered from Facebook users ostensibly for research purposes was ultimately shared with strategic communications firm Cambridge Analytica to develop political campaigns. In another instance, Danish researchers drew fire when they shared a dataset in a web forum for social science researchers containing sensitive information from 70,000 users of an online dating site. And last September, researchers in California came under fire for using 35,000 photos from online dating apps to develop an algorithm that assessed sexual-orientation based on facial features.
听
鈥淭here are things that can be done with this information in the name of science that people might be uncomfortable with, even if it is technically public data,鈥� says Fiesler, who worries such controversies might dampen user trust in a vital area of research, which could reap big social benefits. 鈥淛ust because we can collect data doesn鈥檛 mean that we should be able to use it however we like.鈥�

In September, the National Science Foundation granted Fiesler and colleagues at five other institutions a $3 million grant to begin to develop ethical guidance and standards for the nascent field. She said she hopes her work will inform the way researchers design and conduct studies and also serve as an eye opener to social media users.

鈥淚 would never want my work to suggest to people that they need to stop using social media or lock all their stuff down. But it is important for people to realize that there are other uses of your social media content than what you intend.鈥�