anonymity in research – how now?

Most of us are now very findable online. It is not impossible, once you have someone’s name, to find out where they work and who they work with. Linked in, facebook, twitter, electoral rolls and so on all provide readily available information about people.

Once you know where someone works, a lot more detail comes within reach. Because of this ‘findability’ it’s almost impossible these days for someone who is a practitioner researcher or auto-ethnographer to completely disguise their location and their participants because they themselves are locatable. Teacher-researchers for instance all work at a school which can be found simply via tracking them. Schools generally have websites which often have the names of all staff as well as pictures of students doing things. They put their newsletters on line. It’s not too hard then for someone who is so-minded to pick up a teacher-practitioner thesis, get to the name of the school, identify some of the people involved in the research and possibly even find pictures and names of the staff and students who feature in the thesis as anonymised persona.

I was recently in a viva where one of the examiners did just this online detective work, as a way of raising with the practitioner–researcher the dilemma of whether it was actually possible to promise anonymity. It had taken less than five minutes for this examiner to track down the exact location of the research site and find out the identities of some of the people involved in the research. Now the examiner wasn’t doing this to be nasty or invasive, but to raise the question of how, in the kind of data-dense world in which we now live, it is actually feasible to guarantee anonymity in the way we once did.

The question of identification of course goes beyond practitioner research. We are wrestling with anonymity in one of my current research projects. Because of the specificity of what particular sites offer it won’t be too difficult to work out who some of them are. So we can either:

• name the sites with their permission – but does this mean that there is a temptation to only say ‘the good stuff’ because otherwise it could do harm to their reputations and our relationship with them (also discussed here)? And if we name the sites and programmes is it then possible to promise anonymity to the people we talk while there? Probably not.

• not give much detail about the sites and programmes – this defeats the purpose of doing population-place-conscious case studies which work from the premise that specificity matters, or

• not include the actual case studies in our public report – this means that the kinds of discussion we have about our results and the claims we make won’t be well substantiated.

We are discussing the issue of anonymity with the people involved in our research, but I cant say we’ve reached a simple solution yet.

And these are only two examples of the anonymity questions that arise because of contemporary digital practices and the traceability of data. Assuring confidentiality is tricky for exactly the same set of reasons. Digital theses – and their appendices – can be read by anyone, so research participants, but equally intimate others who were not involved in the research but curious about it, can potentially access data to work out who said what, about what, and who. This may not be a Good Thing but actually do harm.

Because of all this murkiness, there are now serious conversations going on in some parts of the academy about what it means to give consent to be a research participant, whether consent means consent always and forever, whether consent is an event or an ongoing process, and whether consent covers re-use of data. Researchers involved in digital research now routinely talk about the right to have your personal data forgotten.

Questions of anonymity, consent and confidentiality concern us all and they are crucial in the education of the next generation of researchers. I am concerned that very little of this conversation seems to make it into basic doctoral training – or perhaps it’s more a case of these issues randomly making it into courses in some locations some of the time.

I also worry that ethics committees might act conservatively in ‘grey digital areas’; it’s now a case I think of there needing to be iterative temporary ethical settlements that change as we understand the issues. These settlements need to be informed by widespread debate about what it means to be an ethical researcher now. Having hard and fast ethical rules, rooted in historical precedent, is probably not helpful and potentially difficult for all researchers. And difficulties could particularly arise for doctoral researchers who may well have spent the best part of a year planning a research project only to have it rejected because it is on, and/or in, a difficult and shifting digital terrain. Getting an ethical refusal can seriously knock the doctoral researcher off course.

These tricky issues are not going away. They will become more and more tricky the more we amass digital footprints and interlocking and enormous data bases. Yet I still hear ‘doctoral training’ conversations where the notions of anonymity and confidentiality are used as if researchers work in contexts unchanged from when I did my doctorate…

About pat thomson

Pat Thomson is Professor of Education in the School of Education, The University of Nottingham, UK
This entry was posted in ethics and tagged , , , , . Bookmark the permalink.

14 Responses to anonymity in research – how now?

  1. I think the issues raised in your post are increasingly relevant in research and probably under-explored in the literatures — something to write more about, perhaps?

    Reading this reminded me of my long-standing and genuine confusion around what people mean by ‘anonymity’ and ‘confidentiality’ and the difference between the two. As you use both terms in this post, how do you see them? Many thanks..

    Like

    • pat thomson says:

      They are highly interrelated, if not interlocking concepts, but crudely, anonymity is about identity as in I won’t say who you are, use your name, say where you work when I write/speak… and confidentiality is about access, and I won’t allow anyone to know what you said about them or things that concern them in anyway … I agree they are tricky to separate out with a big overlap in the middle. And I think I’ve used confidentiality sloppily in the piece… But I was thinking of appendices in particular.

      Like

  2. Hi Pat,

    Thanks for this post – these matters have been on my mind recently, having just completed a paper with Brigitte on Twitter, where I stumbled upon the ethical difficulties of online data. Rather than the “opportunistic information grab” Helen Nissenbaum talks about here: http://www.mitpressjournals.org/doi/abs/10.1162/DAED_a_00113#.U2IcwfldXTo we opted for a much greater level of anonymity than is usual in social network research http://www.plosone.org/article/info%3Adoi%2F10.1371%2Fjournal.pone.0094785 (Analytical Approach, paragraph 3).

    As a newcomer to social data, I have been surprised that ethical guidelines are not as firm as they might be. Perhaps the temptation of all that data is too great. And, yes, perceptions of the private/public boundary may be changing. However, I think researchers have an obligation to inform and communicate with (often unwitting) participants, rather than sit back and dredge the internet for data. Definitions of ‘research subject’ in some fields may also need revisiting….

    Like

    • pat thomson says:

      I’m a bit worried about an open data move too. There is some precedent in qual research to making transcripts available for example for reuse, as in Bourdieu’ s Weight of the World. I think that’s a much riskier strategy now than it was then.

      Like

  3. This is a really interesting post and very relevant to a research project I’m starting on pedagogy in schools in ‘disadvantaged’ contexts. The teachers aren’t co-researchers in this project but I’ll be feeding findings back to them for discussion and to reflect on potential implications for their practice. I’ll be interviewing teachers and pupils and issues may well arise which could be seen as criticisms of teachers’ practice. My way of pre-empting the anonymity problem is to design the project as a four school case study where I’ll be interviewing 9 pupils in each school. I’m hoping this will help in anonymising and drawing out findings as patterns from across the schools. Although as it’s a small project it does sacrifice some depth in favour of breadth. It still does leave the problem of specific examples and cases which can be very illuminating in qualitative research and it will no doubt be a challenge to balance this with anonymity through making more general points.

    On a wider basis, the issue of promoting the research more publicly through social media in light of anonymity, a question emerges about whether the projects could be traced back to the schools. Something I’ll need to consider carefully – and perhaps in conjunction with the participating schools themselves.

    Like

  4. lizit13 says:

    Thank you for this posting Pat. It is an area that concerned me greatly in my doctoral research. I debated long and hard over whether I could/should legitimately include an auto/biographical narrative in my data set. In thinking this through, I became acutely aware that this was not only an issue for me personally, but that in interviewing parents about incidents concerning their children, they were potentially as recognisable as me. Recently there have been discussions about similar issues on an email list I subscribe to, and at a recent presentation, I heard Tom Couser speaking of a woman who had been shocked to find her case history published in a medical text.
    The whole question of what we mean when we speak of anonymity and confidentiality needs far more unpicking – and we also need to recognise that some people are more than happy to share their data openly, whereas others are in a different position. And whatever the positioning or participants and the settings they live/work in, there are other ethical considerations in respect of our duty of care. But as you say the risk is that ethics committees could be overly cautious, with other attendant problems.
    Tricky!!!

    Like

  5. VioletSky says:

    I wanted to include auto-ethnography as part of my thesis but a/ my supervisor doesn’t see it as a valid research methodology and b/ I was concerned about putting that personal information out there and having people easily being able to connect it to my family, my friends and myself. In the end, to my disappointment, the auto-ethonography section was scrapped. I strongly believe that my personal experience would have leant something important to my thesis but in the end, was too uncomfortable with putting that information out there. I am still extremely disappointed even now.

    Like

  6. Anonymity is a hot topic, even if you “only” work with quantitative data, and esp. when working with companies or small groups. In companies, if participants fill out the surveys at work, they use the company network, so the communication between the survey tool and the work pc better by encrypted. With small groups, even aggregation does not help much, just imagine all members of a team giving low ratings on supervisor performance. Not to mention that as researcher, you should make your data available (at least in psychology), if another researcher asks for it. Sure, you can require confidentiality, but who knows where the data might end up. And given demographic data, it’s possible to identify individuals quite easily. Not the best conditions to get “truthful” information from participants when they fear their words will become known to the “wrong people”.

    When it comes to the changed nature of research in our information age, I would like to point out another issue. Participants can now easily inform themselves about the kind of research scientists do and what they examine. In some cases, you might interview/survey “naive” participants, but if they have time to check, they can pretty much find out what you want to investigate and what your theories are. Even worse, given digitized dissertations/papers and searchable PDFs, they can enter questions from standardized questionnaires and quickly find out which questionnaire you use, what it’s about, and often even how the scores are generated.

    Interesting times for research …

    Like

  7. Privacy is An Illusion says:

    When I was little in the ‘80s, a single mum with two sons rented our spare bungalow. They lived there for a number of years. Well, one of those two sons is now (in)famous. I had a journalist from the Daily Mail ring my mobile to ask me about him (obviously, I did not give him anything).

    He had managed to use my address as a little child to find my mobile phone number 25 years down the track,. Incidentally, the mobile phone number is registered under my married name.

    On this basis of n=1, I would say that anonymity and privacy is an illusion rather than a reality. How could it be a reality in an environment where a random person can find my contact details based on an address when I was less than 10 years old and a different surname?

    Like

    • Hmm, I think a general problem is that most researchers lack a “security mindset” — “looking at the world in terms of attacks and defenses” (to quote a Coursera course about Digital Democracy by Halderman). You need a computer scientist/hacker who specializes in these issues, and even this person might miss some things. And all it takes is one flaw. Still, while ethics and privacy and anonymity and confidentiality are all related and important, I think it’s important to find a healthy balance. After all, the higher the security, the less you can actually work with it. Without downplaying the issue, there has to be a healthy, rational, workable “good enough”.

      BTW, does anyone encrypt the hard drive, e.g., by using FileVault on OS X? It slows down performance (not much), but at least it’s harder to read all the files on the computer if that notebook gets lost or stolen.

      Like

  8. Les Back says:

    Hi Pat, Shamser Sinha and I have been wrestling with the idea of the new informational environment for qualitative research in relation to our Migrant City book. The whole issue of ‘Findability’ has taken on a different kind of twist for us. We have just published a piece that will be in the book. Also, we have a podcast about it here with one of the participants in the study: http://magiclantern.gold.ac.uk/podcasts/sociology/sonicpostcards/sociableSociologyPC10.mp3

    The article is called: Making methods sociable: dialogue, ethics and authorship in qualitative research – it in the journal called Qualitative Research and is available on-line at the moment. Not sure when it’ll make it in to paper.

    Best wishes, Les Back

    Like

  9. Pingback: Twitter as Public Evidence and the Ethics of Twitter Research | Ernesto Priego

  10. Pingback: Impact of Social Sciences – Publicly available data from Twitter is public evidence and does not constitute an “ethical dilemma”.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s