Critical Participatory Data Science

As a visiting scholar at the Science and Justice Research Center, I have been studying critical social science (mainly feminist science and technology studies, critical anthropology, and political ecology).  This has broadened my methodology to include qualitative analysis (particularly case studies and situational analysis, a grounded theory method) and allows me not only to study complex social-ecological systems, but also to investigate the processes and methods we use to understand these systems and ourselves.

Citizen and Community Science

During my postdoctoral work, I’ve led an international group of citizen science practitioners and scholars to publish a co-written paper on the variety and impact of terminology used to describe citizens, scientists, and citizen science.  In general, we observed that no term is appropriate for all situations, and while it is worth connecting with commonly used terms , it is also important to define and use whatever term is most appropriate for the project and participants.

Better Modeling

I also wrote a “modeler’s manifesto” describing more just, epistemologically sound, and trustworthy modeling practices, under review at Environmental Modelling and Software. Some of my key points include the importance of giving adequate context to your modeling, including detailed methods for those wishing to repeat your study and confirm your results, as well as information about the context of the project and choices that were made which may be important for transparency as well as for historians and scholars of science and technology studies.  I also suggest that collaborating with different people and triangulating between different kinds of knowledge are key practices for better modeling, and that when collaborating, attention to ethics and justice become extremely important.  I presented the manifesto at the UC Berkeley Geography’s colloquium, along with a discussion applying it to my participatory modeling processes, described below.

Participatory Data Science in Rural Zimbabwe

ModelPracticeMaiHove

In addition to these more theoretical projects, I have been quite practically engaged in highly successful mapping and modeling practices with my rural Zimbabwean colleagues, work which has facilitated and supported their ability to identify and solve their own problems and make collective local land-use choices which emphasize their own sovereignty.  We have published a paper in Development Engineering describing and reflecting on our mapping teaching, learning, and application (also see this video of me presenting our participatory mapping projects at the Geospatial Analysis for International Development Symposium at UC Berkeley).  Our analysis of our collaborative modeling techniques as an important component of decolonial data infrastructure is currently under review at Environment and Planning E: Nature and Space, and our model is published and peer-reviewed in the model archive of the Network for Computational Modeling in Social and Ecological Sciences. I am currently preparing a paper on the results of the model itself, to complement the paper on the modeling process (see the section on Complex Social-Ecological Systems).

Imaginactivism

I took part in Joan Haran’s “Imaginactivism” workshop at the Science and Justice Research Center during my time as a visiting scholar. I have a deep personal interest in storytelling in many forms, with narrative speculative fiction high among them.  Joan brought together Starhawk and Donna Haraway with a variety of scholars from around Santa Cruz to work on ideas of imagining alternative pasts/presents/futures.  I think of speculative fiction as a kind of qualitative model for exploring alternative ways of being (see Tyler Cowen’s piece “Is a Novel a Model?” for some parallel thinking).  In my view, really excellent speculative fiction is grounded in a solid sense of how things and people and processes work (or worked) in the real world, takes one or more of those and tweaks it/them in a clear fashion, and follows the cascading implications of those changes.  What engages and engrosses me is how people might function (the same or differently), and how underlying principles of nature or physics might function (the same or differently)… but especially when it is grounded in characters who are believable and interesting.  One thing I admire most about Donna Haraway is her optimism about the future, coupled with critical evaluation of the past and present.  While she points out the issues with how science and technology currently function, she also wholeheartedly embraces “speculative fabulation” — beautifully demonstrated in the “Camille stories” chapter of her recent book, “Staying with the Trouble.”  If we can’t imagine worlds differently than what we have now, how will we build and work towards something different from the status quo?  Otherwise, like many predictive statistical models, we will be restricted to the range of realities expressed in our input data.

I was ecstatic and honored to be a part of the workshop, and I wrote this piece as a contribution.  The prompt was “900 words about a place that is particularly meaningful to you. You might conjure up its future, produce a speculative and disruptive history or trace the contours of an alternate present. Who and what have (had) attachments to that place, and how are those attachments bound up in larger networks of interrelationship? Do those attachments open up ways of imagining flourishing cohabitation (however you conceive of that), or do those attachments need to be disconnected and / or reconnected to create spaces of possibility?”

Where are the Missing Coauthors?

Working with Louise Fortmann, Dan Sarna-Wojcicki, and Meg Perret, I helped analyze a database of almost 300 research papers on participatory research in the development literature. My colleagues coded these articles based on their level of participation with community members, whether there was indigenous knowledge or gender-related content, and what the institutional affiliations of the authors were.  We then fit statistical models to find out if any of those factors governed whether academic authors doing self-described ‘participatory’ work either coauthored or acknowledged their community partners.  There were only 16 papers which had community coauthors, and only just over half of the papers acknowledged community contributions to the work. My colleagues interviewed the individuals who did coauthor with community members to find out more about barriers and factors determining coauthorship.  This interdisciplinary project features me as a statistician collaborating from the beginning with sociologists skilled in qualitative analysis and interview methods. This project was published in Rural Sociology (version in French here), and our advice to explicitly consult community collaborators regarding authorship was incorporated into the recent update of the Rural Sociology Society’s Policies and Procedures (see Appendix E: Code of Ethics, section 16(d) ).