Data journeys in a research career #2: Claudia Zucca – Data analysis and the future of social sciences

Claudia Zucca @cla_zu, UK Data Service Data Impact Fellow and Marie Curie Early Stage Researcher working on the VOTEADVICE project, shares her journey about how, with data, it is possible to measure in an efficient and precise way the impact of a policy.


In 1803 Henry de Saint- Simone, introduced the idea that social phenomena could have been studied by using the same theoretical instruments, the same methodologies, used by hard scientists. A little bit later Auguste Comte called it social physics.

I have always thought that this idea is incredibly intriguing and I never understood why we should believe that social phenomena cannot be observed and measured as much as physical phenomena. As a matter of fact, they are not really the same, and for this reason, the name social physics did not live through. There is, in fact, a very clear difference between physical phenomena and social phenomena. Social science phenomena are much harder to observe than physical phenomena. There is an objective physical truth out there that can be observed and measured, collecting as much data as needed. On the other hand, there are several, plenty, of “social truths” out there, each one different from the other according to the peculiar characteristics of the observed community of people. This makes the job of social scientists extremely hard and, probably, for this reason, incredibly fascinating as only challenging things can be.

Social scientists have to set boundaries when they approach a research topic and try to understand the social limits of the phenomenon they are observing. For instance, one cannot observe the impact of family influence in an election in a country with no free elections, as much as one cannot observe the consequences in your education for having your grandparents fighting in the Second World War if your family is from Switzerland since five generations. For this lack of generalisability and the objective difficulties in the collection of the data, several people decided to throw the baby out with the bath water; social phenomena were only suited for description or qualitative evaluations and collecting data about social phenomena was pointless.

The futile methodological debate that is still going on at certain levels wants to characterise qualitative or quantitative social scientists as winners, ignoring the fundamental methodological issues behind this construct. Qualitative and quantitative approaches to research do different things. They ask different questions, provide different answers, from different perspectives and they are therefore rather than competing, complementary. There can be excellent qualitative research and terrible data based research as much as astonishing results coming from a hypothesis testing and weak considerations coming from a discourse analysis.  If no one would ask if a physicist is better than a philosopher of science, why people ask if qualitative research is better than quantitative? The difference is in the researchers and the epistemological foundations that they choose to build upon.

Coming to myself, I was one of those people that never decided if she liked better physics or politics, I loved both very much. Therefore becoming a quantitative political scientist was the clear outcome of my two passions. Working with data for me is much more appealing than any other approach. I like to think that the data are there to tell me a story and I use my statistical tools as a dictionary that allow me to understand the language of the data so I can speak up for them. According to the scientific method, you formulate a hypothesis, and then you question your data. Sometimes they tell you are right, sometimes that you are wrong, and you move on discovering the socially constructed reality step by step, trying to challenge the old paradigmatic truth with new bits of knowledge that enable new paradigms to step up, in a Kuhnian sense.

I also believe that working with data connects researchers’ work with the real life of people in a very efficient and strong way. For instance, if a political actor implements new social policies it is not necessary to try to guess what is going to happen or happened. With data, it is possible to measure in an efficient and precise way the impact of a policy – a powerful instrument to make our everyday life better.

Social sciences are traditionally placed among the humanities subjects, but this is reductive, data analysis is a relatively new field that is becoming day by day more popular among people interested in human interactions. The advent of the internet started to produce enormous quantities of data that are providing easy to collect information about human behaviour as we could have never possibly imagined before. In a couple of years being able to handle data is going to be critical and the level of mathematical sophistication required in addition to the ease of collecting these data will make social sciences much closer to the idea of a social physics. The best is yet to come! Follow Claudia on Twitter @cla_zu.


Leave a Reply

Your email address will not be published. Required fields are marked *