Skip to content

Researcher-imposed context collapse in digital ethnographic research

Author: Christian Ilbury

Please cite as: Ilbury, C. (2022): Researcher-imposed context collapse in digital ethnographic research. In S. Kotilainen (Ed.), Methods in practice: Studying children and youth online (chapter 12). Retrieved DD Month YYYY, from, doi:

Please allow external media cookies to access this content.

Video: Talking abstract - Researcher-imposed context collapse in digital ethnographic research

If you are experiencing issues with the video player, please watch the video here on our YouTube channel. We are in the process of fixing this issue. Please excuse the inconvenience.

Context collapse, the flattening of disparate audiences into one, has been a central concept in theorising social media interaction. An abundance of work has explored how users manage context collapse through a range of self-presentation strategies including censorship, audience modulated styles of interaction, and privacy settings (e.g., Marwick & boyd, 2011; Gil-Lopez et al., 2018). This paper examines the researcher-imposed context collapse that is created in digital ethnographic research based on a year-long blended online/offline ethnographic project in an East London youth group conducted between 2016 and 2017 as a ‘blended’ ethnographic approach (Androutsopoulos, 2008) which explored how individuals’ networked interactions are embedded within broader ‘offline’ contexts of use. I consider here the ethical challenges that arise when conducting research on the digital practices of young people. 

Research which has examined young people’s digital practices has tended to show users design their messages and profiles for an imagined audience that comprises peers and those with similar interests (Marwick & boyd, 2011). As such, digital ethnographic methods such as ‘lurking’ or ‘invisible observation’ (Varis, 2014), impose a context collapse of researcher-participant interactions. 

The main online component of this research involved the adoption of digital ethnographic methods, such as ‘lurking’. In addition, I also collected and analysed a subset of the young people’s social media posts, primarily by accessing a subset of users Snapchat Stories. To generate a corpus of social media posts, the stories were randomly sampled using screen captures2. Herein lies the issue. As discussed previously, young people design social media content for a specific audience – usually their peers. However, by viewing and accessing the young people’s stories, I had, inadvertently, created a type of context-collapse. 

This poses several ethical and practical challenges. In the interests of space, I focus on two main issues here. The first issue is that, given that Snapchat permits users to view those who have accessed the story, the individual could (potentially) adapt their behaviour in response to being observed (i.e., the ‘observer effect’). The second is that, if the individual does not change their habits, the researcher is now privy to the users’ unmodulated styles of online presentation which, as previously discussed, are designed for their peers, not the researcher. 

Whilst these issues may be relevant to other (offline) research settings, they are potentially heightened in digital contexts given that the informal nature of social media environments can lead users to (unintentionally) reveal sensitive or personal information. Indeed, research has shown that digital environments tend to reduce social inhibitions, with individuals disclosing personal information (e.g., Roberts et al., 2000), adopting alternative identities (e.g., Ilbury, 2019), and engaging in miscreant behaviour (e.g., Lane, 2019). In my own research, I observed similar practices. For instance, users would post screenshots of private interactions with friends, upload videos of unruly behaviour, and reference personal details, such as phone numbers. 

Whilst individuals had agreed to participate in the study and contribute to the corpus of social media posts, it would be problematic to assume that all posts were suitable for extraction and/or analysis. As researchers, we have a care of duty to participants – particularly when studying the practices of young people. Thus, I would argue that, following Tagg and colleagues (2016), it is necessary to adopt a reflexive approach to research ethics, where procedures and methods can be adapted in response to the practices and expectations of the participants. 

For this reason, I adopted several practical and procedural measures to ensure that the data sampling procedure was responsible. The first procedure utilises the ‘screenshot notification’ affordance of Snapchat. On the app, it is standard policy for Snapchat to notify to the user that their image or video has been screen captured. Whilst this could be potentially viewed as a limitation of the data collection process, I suggest that this affordance can be used as an additional safeguarding feature. In other words, since participants were alerted to the stories that I sampled, it was possible for the young people to request for the image/video to be removed from the corpus. This was utilised in a few cases where the participant was concerned about the potential for the upload to be misinterpreted by those not privy to the original interaction.   

The second measure that I employed was a reflexive approach to sampling and analysis. Although I had anonymised the data by assigning participants pseudonyms and editing images/videos to remove personal details or identifying information, I extracted and analysed only those stories which, if decontextualised, would not be harmful to my participants. In future, these issues could be further addressed by implementing a debriefing process after the study has concluded where participants could be invited to review their data and if necessary, have content removed from the corpus.  

Lessons learned 

  • the need to adopt a reflexive approach to digital ethnographic research where the researcher adapts their methodological tools to account for the changing expectations of and relationships with participants 

  • it is possible to utilize app affordacnes, such as screenshot notifications, to increase safeguarding measures in our research

  • it is a necessary trade-off to ensure that we are conducting responsible and ethical research on young people’s digital practices. 


1 I gratefully acknowledge the feedback from the reviewers and editors of this volume who helped improve this contribution considerably. This piece is dedicated to the young people who very kindly participated in the study. 

2 Consent was obtained from parents, and assent from the child. Consent for the study was additionally obtained from the youth group management.

Download the full handbook here: PDF.

  1. Androutsopoulos, J. (2008). Potentials and limitations of discourse-centred online ethnography. Language@Internet. 5(8).

  2. Gil-Lopez, T., Shen, C. Benefield, G.A., Palomares, M., Kosinksi, M. & Stillwell, D. (2018). One size fits all: Context collapse, self-presentation strategies and language styles on Facebook. Journal of Computer-Mediated Communication. 23 (3): 127–145.

  3. Ilbury, C. (2019). ‘Sassy Queens’: Stylistic Orthographic Variation in Twitter and the Enregisterment of AAVE. Journal of Sociolinguistics. 24:245–264.

  4. Lane, J. (2019). The Digital Street. Oxford: Oxford University Press.

  5. Marwick, A.E. & boyd, d. (2011). I tweet honestly, I tweet passionately: Twitter users, context collapse, and the imagined audience. New Media & Society. 13 (1): 114-133.

  6. Roberts, L., Smith, L. & Pollock, C. (2000). ‘U r a lot bolder on the net’: Shyness and internet use. In W. R. Crozier (ed.), Shyness: Development, consolidation and change, 121–138. London: Routledge.

  7. Tagg, C., Lyons, A., Hu, R. & Rock, F. (2017). The ethics of digital ethnography in a team project. Applied Linguistics Review. 8 (2-3): 271-292.

  8. Varis, P. (2016). Digital ethnography. In Alexandra Georgakopoulou & Tereza Spilioti (eds.) The Routledge Handbook of Language and Digital Communication, 55-68. Oxford: Routledge.

Cookie preferences

We use cookies on our website. Some of them are essential, while others help us to improve this website and your experience.