Interview: Emmanuel Letouzé, Data-Pop Alliance on Big Data and Human Rights – A Complex Affair
We discuss the founding story of Data-Pop Alliance, the applications and implications of Big Data on Human Rights and the need for penetration of Data Literacy.
AR: Q3. In your 2012 paper "Big Data for Development: Challenges and Opportunities", you mentioned that in order for Big Data to work for development, we need to become "Sophisticated Users of Information". Can you elaborate on that?
EL: As you point out that was in 2012—actually written in 2011—which feels like a very long time ago given the speed at which the space, and I would say my own thinking and ‘understanding’ (i.e. what I think and what I think I understand..), have changed. So what it meant then isn’t exactly what it would mean now. But first the phrase isn’t mine; it is from Jim Fruchterman—which is duly referenced in the paper.
What he meant and what I meant using it was that anyone basing a decision on various pieces of information from different kinds of sources should be aware and wary of their inherent limitations –and those of the analysis and treatment they result from. In his post, Jim Fruchterman pointed to the need for “well-trained and experienced professionals in positions to make these critically important decisions.”
A classic example is the risk of basing humanitarian response in the aftermath of a natural disaster on tweets of SMSs sent and received from affected areas. It’s very
possible that the areas with the most tweets and SMS calling for help may be those least affected, because you can’t send an SMS if you are critically injured or unconscious, and because the poorest and most vulnerable people may not have cellphones to start with. So being a ‘sophisticated user’ takes a combination of technical skills in econometrics, statistics, etc., and also domain and contextual expertise, including the ability and willingness to ask for advice, and common sense.
Three years later, I would stress the importance of being sophisticated commentators and actors of and on information, and data—and what I think is missing in those areas. I think there are lots of simplistic assumptions going on about the nature and role of information and data in societies. Something I think quite strongly that the primary reason behind why the world is in such a bad shape is not lack of information. Poverty is not primarily due to a data deficit, or even a deficit of sophisticated users of information among ‘decision-makers’. I am using the term ‘among decision-makers’ because in the minds of many development experts, UN staff, people who have spent their lives working with governments, ‘decision-makers’ tend to be equated with policymakers, governments, CEOs. There are people who have a very mechanistic view of the world—I call it Bismarckien—and think that all governments and officials are enlightened and well meaning, such that bad outcomes can only result from bad policies resulting themselves from bad or lack of information and/or poor skills. I think it’s missing a lot of what is happening and needs to happen. In a sense the data deluge has given a perfect excuse for the failures of the past: it’s because we didn’t have data!
Data are very powerful tools, and there are a lot about humans and human systems we don’t understand, and Big Data can and I think will fundamentally change our world; but part of the change will involve challenging and probably replacing the old decision-making paradigms, political processes and power structures.
Second part of the interview
Related:
EL: As you point out that was in 2012—actually written in 2011—which feels like a very long time ago given the speed at which the space, and I would say my own thinking and ‘understanding’ (i.e. what I think and what I think I understand..), have changed. So what it meant then isn’t exactly what it would mean now. But first the phrase isn’t mine; it is from Jim Fruchterman—which is duly referenced in the paper.
What he meant and what I meant using it was that anyone basing a decision on various pieces of information from different kinds of sources should be aware and wary of their inherent limitations –and those of the analysis and treatment they result from. In his post, Jim Fruchterman pointed to the need for “well-trained and experienced professionals in positions to make these critically important decisions.” This gets to a critical issue of ‘data literacy’ in the age of Big Data. It’s easy to be fooled by compelling visualizations or fancy models. With lots of data you are always or very often going to find a correlation. But considerations for data quality and representativeness, the importance of distinguishing correlation from causation and the risks of misunderstanding or overlooking spurious correlations, all these basic statistical red flags still apply, or even more so.
A classic example is the risk of basing humanitarian response in the aftermath of a natural disaster on tweets of SMSs sent and received from affected areas. It’s very
possible that the areas with the most tweets and SMS calling for help may be those least affected, because you can’t send an SMS if you are critically injured or unconscious, and because the poorest and most vulnerable people may not have cellphones to start with. So being a ‘sophisticated user’ takes a combination of technical skills in econometrics, statistics, etc., and also domain and contextual expertise, including the ability and willingness to ask for advice, and common sense.
Three years later, I would stress the importance of being sophisticated commentators and actors of and on information, and data—and what I think is missing in those areas. I think there are lots of simplistic assumptions going on about the nature and role of information and data in societies. Something I think quite strongly that the primary reason behind why the world is in such a bad shape is not lack of information. Poverty is not primarily due to a data deficit, or even a deficit of sophisticated users of information among ‘decision-makers’. I am using the term ‘among decision-makers’ because in the minds of many development experts, UN staff, people who have spent their lives working with governments, ‘decision-makers’ tend to be equated with policymakers, governments, CEOs. There are people who have a very mechanistic view of the world—I call it Bismarckien—and think that all governments and officials are enlightened and well meaning, such that bad outcomes can only result from bad policies resulting themselves from bad or lack of information and/or poor skills. I think it’s missing a lot of what is happening and needs to happen. In a sense the data deluge has given a perfect excuse for the failures of the past: it’s because we didn’t have data!
Data are very powerful tools, and there are a lot about humans and human systems we don’t understand, and Big Data can and I think will fundamentally change our world; but part of the change will involve challenging and probably replacing the old decision-making paradigms, political processes and power structures.
Second part of the interview
Related:
