Week 3: (Re)framing Big Data: Activating situated knowledges and a feminist ethics of care in social media research
The audio introduction to this week’s reading can be found here.
I was recently recommended this paper written by Mary Elizabeth Luka and Melanie Millette and found it useful for trying to puzzle through how qualitative research can practically and pragmatically respond to how the digital is shaping the social, without being relegated to the role of critique, relativism and fact checking. The ethical approach to data selection and analysis presented by Luka and Millette got me thinking about how approaches to traditional critical qualitative work can be brought to bear on Internet research.
Luka and Millette use an intersectional feminist lens to think about Big Data. The authors use feminist theorists, Haraway and Arendt, to argue for ethical feminist methods in Big Data research. I was interested to see how feminist theory could practically come to bear on a field dominated by mathematics, algorithms, modeling and visualizing metadata. I am not yet convinced that what the authors present is “practical” but the ethical reasoning they insist Internet researchers, particularly those working with social media, is well worth considering.
Luka and Millette use their own experiences of working in social media research to problematize data as something that can be reduced to zeros and ones in order to develop a representation. The authors remind us that those who select the data, and write the software that scrapes it, all shape the type of data that is collected. Human biases and values are embodied in the decisions made before the data is reduced to something that can be visualized on a scatter chart: “they cannot be an objective trace of a phenomenon, even in social media research where it is tempting to reduce a complex social experience to its digital traces” (p.1).
While qualitative researchers have been making this point for decades, I think it is an important reminder. Internet research can often be dominated by practices and discussions about the ethics of quantitative methods like access to, and the storing of, metadata. Qualitative methods in social media research are lagging but are subject to some of the most interesting ethical discussions. What is a human subject? Do social media users know third parties include research institutes like universities? Do the benefits of researching what people write online outweigh the risks? Luka and Millette enter this discussion by insisting that all methods should be subject to an ethics of care and the ethical work is not yet complete, even if certain approaches have a green light.
The first point the authors make is that no social media research should just be “Big” data because what results is a metaphor for the social that is difficult to shift. Internet research needs to cover multiple registers of analysis: big quantitative sets, smaller sets that can be analysed by hand, thick analysis typically associated with digital ethnography, and lively sets that move with the reality of everyday human activity and the changeable nature of the platforms. The authors propose a feminist ethics of care to frame research practices using social media. This includes the use of different registers of social media data and how researchers practice selecting, collecting, analyzing, maintaining confidentiality, and making data sets open access. Data can be both a means to contribute to a more ethical society, but it can just as easily be violent on the subjects who produced the data, as well as those who analyse it so the authors argue care should be taken at each point in a research project.
Luka and Millette draw on Haraway, and other feminist materialist ideas, to propose speculation as an ethical methodology that considers how the data and the resulting analysis will translate into diverse and power-differentiated communities. Speculation asks the researcher to consider their own values and biases and unpack their own conceptual frameworks and acknowledge how that positioning shapes the data representations they produce. The authors argue that if researchers situate their knowledge in their own power dynamics, it opens opportunities for gaps to be filled.
Luka and Millette then draw on Arendt to propose the building of coalitions to more accurately speculate on the effects of research translation and dissemination, disrupt oppression and power relations, question how the public and private intersect in Internet research, acknowledge costs involved in data production and analysis, and be radically transparent about those costs. The authors suggest a practical approach would be building research tools with participants or bringing participants into the development of funding applications.
The authors conclude by asking their readers ten questions about Internet researcher ethics. They are well worth considering when putting together an Internet research proposal. I have paraphrased them below:
1. What right do we have to research in this space?
2. How are researchers positioning ourselves in this research?
3. Are researchers also participants?
4. Could the research be collaborative?
5. What if participants don’t want to be explicitly involved but still want to collaborate? What are the terms of reference?
6. How can our data collection be visible in ways that acknowledges informed consent beyond third party access being noted in platform terms and conditions?
7. How can debate be advanced about privacy and personal data ownership through this research process?
8. How do we disseminate results in a respectful way?
9. What if shared involvement and collaboration does not improve our research? What do researchers do if the data becomes violent?
10. How will the research contribute to a more ethical world?
Maybe you would like to answer a few of them about your research in the comments. I’d love to hear about how you have ethically thought through your Internet research.
