By Anna Shams Ili, Junior Researcher

When researching digital phenomena, digital ethnography is a useful tool for immersion and understanding the material beyond data points or quantification. As with traditional ethnography, digital ethnography simultaneously opens up for the possibility to experiment with different ways of conducting ethnographic research. In looking at “internet ethnography”, or netnography, general ethical anthropological concerns become visible especially when studying “material” that also frequently contains details of the personal and vulnerable inner life of, often anonymous, users. 

I have spent the past year looking at pro-anorexia content on TikTok as part of the Junior Researcher Programme at ETHOS. Last semester I primarily worked on methods for data collection using network analysis, but even before I began scraping I had started an ethnographic process of immersing myself in pro-anorexia spaces to understand terminology and general themes. For my last post for ETHOS, I wanted to delve into my considerations with netnography more. Through my immersion, my feelings have ranged from primarily ethical concerns towards vulnerable users, to a general frustration about the type of spaces I was in, and to grappling with the understanding that fluidity in especially newer online spaces makes true “immersion” near-impossible. 

A brief overview of digital ethnography and netnography

Digital ethnography most closely draws on and resembles traditional ethnographic methods, by applying those on virtual sites, such as social media platforms or web forums. Pink et al.[1] outline several concepts, amongst those emphasising an experimental form of approaching research and research notes, as well as decentering technology

In contrast, netnography is defined by Kozinets[2] as a different tool. Whilst still drawing on ethnography, Kozinets very much centers technology and digital methods as part of the process. This includes, but is not limited to visual network analysis, which uses sociological network theory to map out actors or concepts as nodes connected to edges. The idea is to use ethnographic methods to analyse and interpret the more qualitative or technical results, thereby “making sense” of digital trace data. 

Situating my project 

This project looks at pro-anorexia, or pro-ana, content on TikTok and how femininity is constructed within this space. For my last post, I delved into the use of network analysis which was also my sampling strategy, using an initial seed list to find co-occurring hashtags that were scraped. 

There are many online forums and general spaces that discuss pro-anorexia – some are ephemeral, there are Discord forums with multiple back-ups and users bounce between them from one link to another. Others are self-hosted web forums that rely on strict guidelines to contain and foster the community. My initial ethnography drew on a web forum that had no entry requirements – posts were available online and didn’t require intervention such as posting anything yourself. This seemed like the best process to start, especially as I wasn’t specifically interested in the individual users but more so the general emerging themes and TikToks they might link to. 

After scraping these with my TikTok profile, made specifically for this project, I quickly realised the potential in using the “For You” page (FYP) to continue immersing myself and to understand what I was working with. And I’m glad I did, as many videos have shown up showcasing a variety of trends and discourses, as well as how many videos are uncaptioned and thereby would have been hidden from my scraping process. 

Alternative field notes

Sitting with several hours of footage of me scrolling my FYP showing one video after another related to bad mental health and eating disorders did make me feel like I was immersing myself into something, but what this something was exactly or how to untangle it analytically still felt illusive. I had written some field notes on interesting forum posts, and saved a few videos I found interesting, but overall I was sifting through things and not making any sense of them in a broader sense. Every post was a tree, and I could not see the forest. 

Returning to other digital ethnography and the idea of unconventional research, I used the ETHOS Play Fair to get a fresh pair of eyes on the material. I wrote some fake forum posts that vaguely encompassed the themes and tones of posts I had seen and written more traditional field notes on. To push people in a different direction (and because more written field notes were not needed), I made them pull a random prompt and either have to draw a field note or make one out of pipe cleaners. 

The field notes themselves were not what you might find in a traditional research paper. But in the reflection notes, participants explained the thought processes behind it – they might not have been specifically meaningful for a wider research community, but they were subjectively meaningful to the person making them. 

One participant wrote in their reflections, on a post discussing media and thinspiration[3]: “The exercise makes me wonder: What other worlds are possible, and what media will make them desirable?”

By having to create something as opposed to simply describing the post, participants were forced to think in much larger terms. It was a big ask for people who were entirely new to the themes and ideas, but as an experiment, it did show the possibility of rethinking material by getting away from describing it. 

Through the reflection notes, I was also able to consider the broad spectrum of feelings people felt, which mirrored my own. One idea I’ve been thinking about especially is the idea of recognition and empathy – that you can empathise with certain feelings and experiences that are much broader than the specific contexts of the eating disorder. It also made it clear to me why immersion is a vital part of the research, especially when working with digital data that could end up just being nodes in a network. Researching these digital worlds without engaging with the content itself would be a disservice to anyone participating, and instead, be biased towards assumptions of what you expect to feel. 

Final remarks on immersion and empathy

While my project does use digital methods as its main methodology, digital ethnography and the idea of decentering technology still allows one to think more critically about the structures that are uncovered, which in my own and the Play Fair participants’ notes have much more to do with empathy and emotion than anything inherently technological. The thoughts and feelings uncovered in the forums, as well as on the TikTok FYP, do not exist due to technology, nor do you need technology to be able to recognise them.

Creating the tasks also started new thoughts: What concepts and ideas did I feel worthy of being reproduced to be read by participants? Why did I pick these specific themes? In my task, participants are reading, reacting, and understanding a single note (or in some cases two), thereby interpreting them in isolation; but it’s only by putting together the posts and reflection notes that I have been able to properly participate in “meaning-making”. To me it felt like they were working on a tiny part of a big puzzle, yet a lot of the posts elicited similar reactions and the participant notes reflected my own thoughts when curating and writing the posts. 

I’ll save the analysis for my thesis, but the Play Fair project and reflections have made me rethink the idea of “immersion” when the field is so fluid. The videos within it will make sense together, but even then the specific sequence of videos is (presumably) unique and momentary, and my reaction and reflections on them are almost impossible to untangle from my emotions towards the content. It might sound like I’m disparaging ethnography on digital platforms now, but I really think the elicited feelings are part of the experience of the FYP as a whole – including the moments of fatigue towards the end of a scroll where I temporarily turn into a concerned parent or clueless U.S senator: let’s just ban all of it. And while my immersion may only touch on one type of algorithm during one space of time, the single puzzle piece is still an important start that can at least hint at the full picture.

Thank you to everyone who participated in the JR Play Fair and contributed to the reflections and field notes. 

[1] Pink, S., Horst, H., Postill, J., Hjorth, L., Lewis, T., and Tacchi, J. (2015) Digital Ethnography 
[2] Kozinets, R. V. (2019) Netnography
[3] Media (usually visual) that “inspire” users to stay or become thin