Q & A with Jana Schaich Borg on Duke, Data, and MIDS
Q. You work with both human and animal participants to study how and why we make social and moral decisions. What led you to this type of research and what do you hope to achieve?
A. Almost every researcher has something in their head or their life that gets them out of bed every morning and motivates them to tackle the hard challenges they’ve voluntarily put on their plate. For me, that thing is violence. It is very difficult for me to get my head wrapped around the fact that humans intentionally hurt each other. The only way I could handle learning about such events is if I tried to do something to stop them or at least understand them. The most effective approach to answer these questions was to try to understand what happens in our brain when we make decisions that affect others. In other words, what would we have to tickle in our brains in order to make us refrain from violence, and more generally, just treat each other better? That’s the question that drives all of the research I do. And these days I’m trying to figure out not only what parts of the brain I would have to tickle, but also how I could learn to tickle them without putting anything through our skull.
As for why I work with both human and animal participants, tackling a phenomenon like human violence requires deep and simultaneous appreciation of behavior and mechanisms. We need to understand the details and complexity of how humans behave, but we also need to understand the mechanisms the brain uses to create such complex behaviors if we are ever going to have any real hope of learning how to change those behaviors. Human behavior is, of course, best studied in humans, but brain mechanisms are most efficiently studied in non-human model systems. My challenge is to identify and integrate the aspects of human behavior and rodent mechanistic neuroscience that are necessary to efficiently change the way we interact with one another.
Q. You are a new SSRI faculty member. How does your research fit with the Institute’s interdisciplinary approach?
A. My research requires deep expertise in neuroscience, psychology, statistics, engineering, and now, programming and data management. My research can’t get done by me alone. On the flip side, I’d like to think—very humbly—that my research can’t be done without someone with my specific background either. When I heard about SSRI’s vision, I immediately thought, “Wow, it would be so exciting to be in that environment and to help realize that mission!” I am biased, of course, but I love the idea that Duke has created a place where important problems that are too big to be tamed by one discipline or department alone can be solved.
I think I also fit into the Institute’s interdisciplinary approach in two other meaningful ways: my work serves as a kind of bridge between the social sciences and data science on the one hand, and between the social sciences and the humanities on the other. By being part of the new Master’s in Interdisciplinary Data Science (MIDS) program, Duke Institute for Brain Sciences, and Kenan Institute for Ethics, I hope to help clear the way for other researchers and students to walk between the social sciences, natural sciences, computational sciences, and humanities more freely.
I’m also deeply connected with SSRI’s mission to build the foundation for social science to engage with a much broader and more interdisciplinary landscape. Much of that foundation will be in the form of providing the data and computational architecture necessary for all disciplines to become more data literate. Another part of that foundation will be in helping established researchers learn how to work on teams with other established researchers. Yet another part will be creating mechanisms for academia to evaluate and give credit for interdisciplinary work. Alongside my own research program, I am hoping to help SSRI build out these pillars of its foundation as we all navigate the opportunity to incorporate Big Data into the social sciences.
I believe the future of social science will be to find insights through Big Data, but to do so in a way that is grounded in the thoughtful traditions of the humanities. So much of Big Data is inherently social. Huge chunks of it are related to social media or are collected and presented in social contexts. Social scientists need to be intimately involved in strategies to collect and interpret these types of data. In the process of doing so, social science will be revolutionized by the new insights data of this kind can bring. In return, I think social scientists will help society find deeper significance in what all these data mean by bringing the traditions of the humanities to the table. In addition, I’m hoping social scientists can take the lead in reminding data-heads of the ethical implications of our actions and choices. We all agree that data is changing society as we know it—we need people who will make sure data is changing society in a beneficial way.
Q. Your Coursera Course, Mastering Data Analysis, is one of the most popular courses on the site. What do you think has contributed to the appeal and success of
A. There is an ever-growing number of people who want to learn how to work with data. We felt an almost moral obligation to try to give people of any age and of any background a way to learn these data skills that would be so valuable to them. That’s the reason the courses were so successful—I think they just filled a true need.
That said, one thing that might differentiate Daniel Egger’s (co-instructor) and my approach from other instructors’ approaches is that we really wanted every student to feel like they could do this stuff. A lot of people get scared away from technical and quantitative areas because they worry they aren’t smart enough or because, at the very least, they worry they will feel stupid or be made fun of. Learning programming, math, and statistics is not about how smart you are, it’s about having a growth mindset, cherishing and learning from mistakes, and trying different ways of learning information until you find a way that works for you. We really tried to help our students feel like we are all in this together.
Q. What do you think makes a successful data analyst or data scientist?
A. In my opinion, there are five main characteristics that define an effective data scientist:
- An ability to communicate effectively.
- An ability to ask and answer the right questions. In practice, that means an ability to translate the problem that needs to be solved or the question that needs to be answered into an appropriate, explicit plan about what data are needed and exactly what you are going to do with those data.
- An ability to think very logically, methodically, and impartially.
- Of course, a good technical skillset with the appropriate quantitative knowledge.
- An insatiable desire to learn and figure things out accompanied by a playful appreciation of failure.
Q. The Master of Interdisciplinary Data Science (MIDS) has recently been approved with plans to enroll students in fall 2018. As the Associate Faculty Director of MIDS, what do you think makes this program different from similar data science programs being offered?
A. Data and data-savviness can help all sectors of life. We want to empower our students with the data skills and rigorous scientific thinking that will allow them to solve whatever problems they are most passionate about, whether those problems are in business, medicine, government, the humanities, or the basic sciences. In addition to the technical and quantitative skills all data science programs have, our MIDS program will put a particularly strong emphasis on:
- Training data scientists who can solve problems in any domain. Students will have the opportunity to work with many types of data from projects inside and outside of the University, and will be taught skills that can be applied effectively to any of these types of data and that will allow them to get up to speed in a new subject matter efficiently.
- Training our students to be deep critical thinkers. When our students are done with the program, they will really understand what it means to be a “scientist.” They will know how to think through a problem logically and methodically, and how to use data and experiments to guide their conclusions.
- The importance of “who cares?”, “what does this mean?”, and “does this solve the real problem?” We intend to train our students to be driven by questions and making a measurable difference rather than by the methods they happen to know at the time.
- Communication and teamwork skills. We believe data scientists will only reach their full potential when they can listen to others, work cooperatively with others, and communicate to others what they have done. These skills are often left out of quantitative graduate programs, but we intend to make them a major component of the skillset our students master.