Northernlands 2 - Four principles for a responsible data collaboration
This transcript comes from the captions associated with the video above. It is "as spoken".
Good afternoon everyone. My name is Stefania Milan and I work as
associate professor media and digital cultures at the
University of Amsterdam. Today I have come to you from this
beautiful garden and in preparing this talk I thought
I'd rather share with you some nice flowers and greenery
instead of yet another white wall that has been accompanied
most of our many, many conference calls and video
presentations over the last few
months. So I really apologize in advance for any occasional bird
or dog or car passing by that might somewhat hinder the audio
of this presentation.
I'm a sociologist, by formation. I've been doing quite
some work at the intersection, sociology, political science,
media studies, and data science, and mostly rather
interdisciplinary and cross disciplinary work. I'm very
passionate in particular about what people do with technology.
I consider data as yet another type of meaningful technology.
Over the last couple of months I've dedicated myself in
particular at understanding how the Covid Pandemic is narrated
through data with a specific focus on those communities and
individuals that have been left at the margins.
So I'm thinking about those that work, for example in
informal economy. People with disabilities but also sex
workers, poor families, deprived children, you name it. The many
precarious and gig workers around our communities both in
the global North and in the global South. What I would like
to share with you today however are some reflections on how do
we try to foster meaningful data collaborations and data
collaborations that become
precious. Not only to us, the researchers. Not only to the
funders, but are directly shaped and become useful also to the
communities on the ground. So to the people and those practices
that we are directly observing.
Today, in particular, I would like to share with you some
reflections. That I did a few years back and not interested in
the context of data-fication of data-fied society
With a co-author on mine, Ciara Milan, currently Marie Curie,
fellow at the University of Gratz in Austria.
With Ciara, we observe comparing our field notes in various
fields that very often
field work on the ground yields very very interesting theory
development. Smashing datasets. Often we do not make apparent
how the data were collected and not just from the methodological
point of view, but from the epistemological and ontological
POV. And in particular we fail to make clear how we develop
relationships with our research subject - with those that are
being researched. For us, and this is the departure of our
reflections, communities are not just research objects. Research
subjects if you prefer. They are rather active agents with their
own values, preferences, visions, strategies, their own
modes and languages of engagement, their very own
specific needs. So how can we make sure we expect then, that
we do not, as researchers overly shape the research in a way that
just benefits us. How can we repurpose the goal and the
sometimes very professional language of researchers. The
professional language of the founders of policymakers in a
way that the research becomes accessible also to those that are
being researched. How do we share our resources? How do we very
importantly consider the communities on the ground? The
individuals on the ground that participate in our research as
valuable resources and consider
also the effort that they make
to collaborate with us and how can we built in the research
process the data collaboration in this case, some operational
mechanisms that warrant that we listen and are accountable to
the communities on the ground. In these reflections were very much
inspired by the amazing work of a urban sociologist from the
United States of America. Boston in particular, Charlotte Ryan.
Charlotte has been working for many, many years with
communities on the ground. The United States.
And at the borders between the United States and Mexico,
as well as in Mexico and other Latin American countries,
in a very seminal piece of work
co-authored with one of these activists that she collaborated
with Karen Jeffrey - a Union organizer - they reflect on
treating or understanding moments and communities on the
ground as skilled learners. So as agents as individuals as groups
that have the capacity themselves on making sense of
their activities and reflect on them. And in this way she
encouraged us researchers. They encourage us researchers to
leverage this ability of communities on the ground to
learn from their own practices
and reflect on said practice. It's in this framework that Ciara and
I came up with a very simple framework for research called
the strap framework. In nature I mean not in this nature, but in
reality a strap is a piece of clothing usually that connects
two items to parts of a bag of a bag or up to parts of a jacket
or a bag to another bag or you name it.
A Strap is also very much
integral to the objects it tries to connect, the objects
it belongs to.
So strap for us in this case, in the research process, stays for
it's an acronym and I'm gonna tell you in a minute what this
acronym hides. But we really want to use the analogy of the
strap in real word to connect
to research and try to make the case that is very important
to building these mechanisms that we offer to
your perusal in order to ensure that data collaborations are
meaningful to both sides of the equation. So strap stays for the
The 'S' stays for sharing.
The 'T' stays for translation.
They 'R' indicates relevance. The 'A' stays for accountability
And finally 'P' is for power. So I repeat, sharing,
translation, relevance, accountability and power.
So let's now go briefly over these five different items that often
hide more complex items. So let's start from sharing.
Sharing is fairly intuitive, right. I as a researcher as the
initiator of data collaboration I might want to share my
findings. Very often however this sharing is not even written in
our research project. It's taken for granted or is simply not
really practiced because sharing entails relation building. Data
collection as well very often entails relation building or a
list. Good data collection
hides very careful relation building. However, we tend to
really underscore this part because this process
are very time consuming.
They often come with the very end of the research project or
research process. When we are ready in the publication phase,
or very simply, we.
Sorry about this mosquito. Very often we simply equate sharing
with sharing simply the publication of final report. A
final academic article. However, this is really not good enough
we argue. For a good data collaboration sharing entails
or has to entail also transparency.
What do we mean by transparency? Way to make the
research process transparent to the research subjects and we
have to share this project from the early on. This connects to the
R letter is going to come up next
But sharing really means trying to understand in the
various phases of the research projects the various phases of
the research collaboration, when can we involve the
research subjects. Also, for example in determining what are
the needs, the desires, and the preferences of the subjects.
But also for example, their research questions.
So we have seen the 'S' of sharing now it's time to briefly
look at the 'T' of STRAP - the second letter. The 'T' stays
in for translation. What do I mean by translation? Well,
translation might entail sometimes translating between
different language between Dutch, Italian, Dutch and
English. But what we really try to signal here is much, much
deeper and goes to the bottom of relation building. Translation
indicates really the conversion, the active effort of converting
my agenda into the agenda of someone else. My language, which
is often a disciplinary language, a professional language, into the
language of the research subject.
But it is not a simply a Mechanical Act. It really
speaks to relation building once more and it points to a very
very subtable but very important distinction that we would like
to introduce here, which is the very key distinction between
research with and research about research about is often what we
practice. Why? Because it is we want to research.
You want to know more about a specific community for example.
I have that the research about supposes some distance.
Entails you know, completely
detachment scientific. Very typical scientific objectivity.
That, you know, make sure that I am the, objective
observer. The distant observer, the expert observer. Now we know
very well, and I'm not arguing against this. This is very
important, very important element of scientific process
- scientific inquiry. However for meaningful data
collaboration to happen and with data this is very specific
need. Because data
opens up before it an immense amount of possibilities of
researching with citizens research subjects. It's very, very
important to go to the bottom of the relationship that tries to
involve the citizens in the process. So here we argue really
with a 'T' of translation for moving away from researching
mainly about into researching with. This researching with
might not concern every phase of research project. Might not entail
or concern every aspect of a
research question. Or very complex research project. But we
do argue that this is very important. It's very important
to at least sometimes for some parts of research project to
build in this relation that is different. That gives agency
back and power back to those that are actively being
researched by us. I give you an example, there's a very
interesting group of what we may call data activist in Milan in
northern Italy that are involving citizens in data
collection about a pollutant. About polluting agents released
into the air by, for example,
you know cars or central heating and other
polluting forces. And what they're trying to do is they not
only give citizens a little device to install, you know, in
their terrace out of their window in the garden, but they actually
involve the citizens from early on in building this very simple
devices. Although wait for those interested citizens that are
interested in able to participate to involving them
into the data analysis, such introducing them to some simple
mechanisms of data analysis and really giving them
power into shaping the research relationship and the research
project. The project I'm talking
about is done by the Off Topic Lab in Milan, but there
are many, many others of this kind throughout the word in
the global South and in the industrialized world.
Now let's move on to the 3rd letter of the strap.
Relevance. We do have to make sure we argue that our research
is relevant not only to us as your developers as data analyst,
as policymakers; those on the powerful side of the equation,
those with the expertise at least in data collection and analysis,
but also to those that are being researched. This entails
starting from and developing mechanisms and research methods
to involve in the research process from day one.
Even from day zero, really.
From the very same research idea that we start with
citizens and people on the ground
Interestingly, a few years back at the University of
Amsterdam, we also did a couple of events where we
involved a civil society organization from the Amsterdam
area in particular into what we call data for the social good
experiments. So we played around with an innovative methodology
that started from the needs so
started from the needs of the communities on the ground from
the society organizations and we tried to meet midway in making
our research questions intersect their research question and needs.
And it was a very interesting form of dialogue, very similar
to what Charlotte Ryan and Karen Jeffrey experimented in the US
for the last 20 or plus years
and I'm sure there are many, many other
groups that practice... the effort of making data collaborations
relevant also to research subjects, so making sure that our
research questions - at least some of them - get closer to what
those who are being researched also care about.
And we move now to the 4th letter, in STRAP, the 'A'. The A
stays for a very, very informal, important component of any
research project: accountability. Now very often as researchers,
we understand accountability
towards our employers, but more importantly towards
our funders, right? We spend an enormous amount of
time with very often very tedious, but nonetheless very
important reports back to the funders, where we explain
very carefully how the money, was accurately spent, and how
we try to make the research process cost efficient. How we
really did the best possible choice given the situation.
But we don't spend enough time and enough energy thinking
about what we consider a very important aspect
of the research project of meaningful data collaborations,
but also meaningful research relations, which is the
accountability towards our research subjects. It's similar to
...it ties into what we named translation. And it concerns
being accountable to the research subjects, so making
sure, for example, that we try not to harm very, very
important. Imperative that we derive from example from
the hacker culture but also from many other cultures as well. From,
for example, the culture of the Medical Sciences, but also from
the very important need of making sure that we are
accountable, not just at the very end of the project again,
but while the project.
For example. We should ask interrogate ourselves,
especially we do research with people on the ground that might
be, for example, human rights defenders or people that would
research or activities with communities that are potentially
harmed or endangered.
You know, by the state, by potential other enemies.
Then we make sure that we report back to them.
We involve them in the process, also assessing the
risks and this is what we mean by accountability. Assess the
risks of research projects throughout, not only the costs.
Financial cost if you want.
the risks in terms of for example repression or the risks
of exposing certain practices or you name it according to what is
your specific research situation of data collaboration that you
might have in mind. So what is very important is that this
becomes an activity of building bridges and that these bridges
across very time very often throughout the research
projects and that we disclosed what also our risk assessment is
because sometimes the enthusiasm over shadows
the risks also for the people in the ground.
And finally we get to the last letter of our STRAP framework,
the 'P'. 'P' stays in for power. What do we
mean by power?
Well, you know, often we are the power side of
the equation. We are those with research money with a research
grant to observe people on the ground. Sometimes the power is
not even, you know financial power or anything, simply the
expertise and ability to make
time. Make room for observing practices that people actually
lived in or spent their life fighting for. And you know, we
have the distance that we have the time with the resources that
allows us to do so compared to the people on the ground.
And this, you know, this really brings me to discuss power.
This expertise at this time. This money sometimes
results in very severe power imbalances between us and
those that we are researching. It's very important to get this
power... although very often we cannot really offset it...
To give this power adequate consideration and try to, in
a way or another, act against it or share resources if we can.
But in case of not being able to do so, at least acknowledging it.
Not just, you know, in the acknowledgement section of an
academic paper, but especially
when it comes to also designing and implementing
the more research project.
So to sum up.
The STRAP framework
Stands in for sharing, translation, relevance,
accountability and power.
We ask you to keep it into account as a source ready to use
and simple recipe to make sure that our research project and our
data collaboration is able to create many full connections
between the researchers and those being researched while
embedding the research process into an ongoing process or a
wolpe process of social change. Thank you for the attention.
Have a good rest of the day.
Principal Invesitgator, DATACTIVE
Stefania Milan is Associate Professor of New Media and Digital Culture at the Department of Media Studies, University of Amsterdam. Her work explores the interplay between digital technology, activism and governance. She is the Principal Investigator of DATACTIVE, a project financed by the European Research Council exploring data- and algorithmic-mediated forms of political participation (data-activism.net). She is also the Project Leader of "Citizenship and standard-setting in digital networks", funded by the Dutch Research Council, and co-Principal Investigator in the Marie Curie Innovative Training Network "Early language development in the digital age".
In 2018-20 she directed the Algorithms Exposed (ALEX) project, tasked with developing open source software tools for auditing personalization algorithms on social media and online shopping platforms. In 2017, she co-founded the Big Data from the South Research Initiative, a network of academics and practitioners critically investigating the impact of datafication and surveillance on communities at the margins. Stefania holds a PhD in Political and Social Science from the European University Institute. Prior to joining the University of Amsterdam, she worked at, among others, the Citizen Lab at the University of Toronto and the Central European University. Stefania is the author of Social Movements and Their Technologies: Wiring Social Change (Palgrave Macmillan, 2013/2016) and co-author of Media/Society (Sage, 2011). She enjoys experimenting with digital and action-oriented research methods and finding ways to bridge research with policy and action.
Nothernlands 2 is a collaboration between ODI Leeds and The Kingdom of the Netherlands, the start of activity to create, support, and amplify the cultural links between The Netherlands and the North of England. It is with their generous and vigourous support, and the support of other energetic organisations, that Northernlands can be delivered.