Unit III: Ethical Research Conduct [Reading time: 10 Minutes]

As mentioned in the previous unit, I used TAGS to scrape data for a research project in late 2019. Over the span of approximately 5 months, I collected/half a million tweets. Pretty early on in my Twitter scraping, I realized that I was getting more information than I had asked for or expected. For instance, I was only interested in knowing what kinds of tweets were using hashtags and who was tweeting them. However, the Twitter API gave me information on the user profile location, the number of retweets, Twitter ID, replies, mentions, media, and profile image.

I was struck by how much more data I was able to get. As a graduate student who was just beginning their data collection, I was elated. This was a boon to put it simply. I was able to get a lot more data without having to do a lot of work. At the time I didn’t really think about the power implications or ethics behind that. It wasn’t until 2021 when I started doing participant observation on Twitter, that I started to reflect on my existing Twitter dataset. The question that I had for myself was how private is social media data?

And here I am indebted to Internet and social media scholars such as Christian Fuchs, Matthew Williams, and Moya Z Bailey, who have not only written extensively but also offered us models on how to conduct social media research in an ethical and transparent manner, especially when working with vulnerable communities or online hate groups. Their scholarship served as a guide for my own work as they alerted me to how desperately social media researchers need to be thinking about questions of ethical conduct, regardless of how tedious or arduous it may appear.

Issues of Privacy

• Commercialism: Twitter is a public platform that is driven by commercial interest. Their guidelines on privacy should not be a starting point for researchers and academics. Instead, we researchers should think about the implications behind quoting a tweet verbatim in our study, even if it is coming from a relatively unknown, small Twitter account.

• Public data: Even if Twitter is saying that its platform is public, that does not mean that its users are aware or even onboard with how their tweets will be used outside of the platform. Just stating that a dataset is public, does not mean its contents are always public, especially if the information that we have collected about a group of users has not consented to us using their words for a research study.

• Lastly, anytime you use social media data for academic purposes, it always has the potential to release sensitive information about people. What I mean is that we need to consider what constitutes sensitive information in the context of our individual social media research. What I, as a researcher, consider to be sensitive information may be totally different or at odds with what the users who make up my data, consider to be sensitive information.

So, how do we ethically go about conducting research using social media data?

The existing scholarship on internet research ethics suggests that we create context-specific ethical frameworks for our research methodologies. To do so, we can implement a three-step process. The first two steps are research-driven. They involve us reading up on existing models of internet research ethics, and getting acquainted with data privacy laws from the region we are conducting our research from. The last step is to create a model of our own, that is specifically applicable to our research context.