“I am interested in promoting diversity in the Statistics, Data Science and R communities”: An interview with Isabella Gollini

Isabella Gollini is Assistant Professor of Statistics at University College Dublin where she joined in September of this year, having studied for her PhD there. Prior to her appointment at Dublin, she was Lecturer in Statistics at Birkbeck, University College London and a Research Assistant in Bayesian Statistics at University of Bristol.

Her research activity aims to integrate statistical methods with various applications through cross-disciplinary collaborations. As an applied statistician, Professor Gollini has been working closely with scientists with different expertise (e.g. engineers, geographers, computer scientists, etc.) in order to yield important practical results as well as developments of new statistical techniques.

Alison Oliver spoke to Professor Gollini during uSER! 2017 in Brussels where she was a guest speaker.

1. What was it that first introduced you to statistics as a discipline and led you to choose to study the subject at university?

I have always enjoyed Mathematics. In Italy, unfortunately, we generally don’t study Statistics in high school, and we therefore don’t have many ideas of what statistics really is. However, I soon came to know that there was this very new degree program in Mathematical and Computational Statistics in Genoa, my hometown in Italy, so I decided to investigate further that opportunity. It was explained to me that there were very high chances to have a job immediately after the BSc so I decided to give it a go. There was also the opportunity to have a French-Italian double degree and that was a big plus. I have to say that I enjoyed it so much that I decided to do a MSc and then a PhD.

2. After studying for your MSc in Statistics at the University of Bologna, you then headed to University College Dublin for your PhD? What led to this opportunity to study abroad?

I’d been interested to study abroad since my undergraduate experience in France. So, I took the chance of visiting ESSEC Paris for six months in order to prepare for my MSc thesis. Everything seemed to point in the right direction and it was a great opportunity for me. So, after my MSc studies I decided to explore the possibility of a PhD in either UK or Ireland. UCD looked like a very natural choice to pursue research in cutting-edge fields, and learning at a prestigious research university.

3. After teaching at Dublin, NUI Maynooth and Bristol, you joined the Statistics department at Birkbeck, University of London in 2015. How are you finding relatively new experience?

I really liked the unique environment at Birkbeck: a stimulating and supporting institution from a research viewpoint and very different with the respect to other academic places from a teaching viewpoint. In fact, I had to adapt my teaching style to best fit Birkbeck students who are mainly mature and/or working students. I learned the value of an interactive classroom where students are active in investigating how to connect their everyday problems at work to what they study during the classes. Even if I lectured one of the introductory modules on the probability and statistics, students could already see how important statistics is for their jobs and real world impact. And I must say that was very rewarding. From September this year, I have taken up a new position back at University College Dublin. Whilst I was sorry to leave Birkbeck, I was very happy to join a very well-known department and I had finally the opportunity to live in the same city as my husband!

4. What do you think the most important recent developments in the field have been? What do you think will be the most exciting and productive areas of research in statistics during the next few years?

Thanks to recent advances in computational statistics, we are now going to be able to infer very large and complex datasets. The possibility to perform large numbers of computations has opened many important new possibilities for Statistics and Machine Learning. For this reason, it is very important to teach students about the advantages (and limitations) of these methods. In fact, these advances often need the development of new methods and supporting theories. So, in the next few years, there is going to be the need of carrying out research on both computational methods and theory.

5. Your research has been published in journals and books: is there a particular article that you are most proud of?

My first article was about proposing mixtures of latent trait analyzers for the analysis of binary data and I’m proud of that because I almost completed it by the end of my first year as a PhD student and that represented a very important achievement to me. I think it is a very complete paper as it features a new modelling approach together with model selection procedures and very efficient computational methods to deal with large data sets.

6. You are a mentor at R-Ladies and the teaching leader at Forwards, the R Foundation taskforce on women and other under-represented groups. Can you tell us more about these roles? What do you do to involve women and bring them into the spotlight?

I am interested in promoting diversity in the Statistics, Data Science and R communities. I think it is fundamental to develop new teaching methods and tools matching the interests of a wider community that is not just made of the white men, and therefore I started my role as teaching leader at Forwards the R Foundation taskforce on women and other under-represented groups. R-Ladies is another great community which promotes a welcoming and safe environment for women interested in R. What is great about R-Ladies is that every woman can be involved in different ways depending on her skills and cultural background. Every time I organise a course or a session in a conference I always do my best to invite a diverse set of speakers and encourage a diverse participation. When teaching, I try to incorporate as many real-life and everyday examples as possible to continue to encourage women and diverse students to stick with the challenging world of Statistics and Data Science.

7. What has been the best book in statistics you have ever read?

Well, there are several nice books that I read over the last decade. During my first years as a PhD student, I used a lot Pattern Recognition and Machine Learning by C. Bishop. I’ve found it extremely accessible and useful. I would suggest this book to beginners who want to understand the Bayesian perspective on Machine Learning and approximate inferential methods.

8. What would you recommend to young people who want to start a career in statistics?

I think my simple answer is “just do it” because statistics can open up many opportunities! Statistics and data science provide people with critical thinking skills to evaluate, analyze and solve real-world problems in a creative and innovative way. Sometimes people start to be interested because of applications to economics, social science, biology, etc. Then some people decide that they don’t like working in a specific area but they would prefer to apply statistical skills to other areas. This is possible because statistical skills are highly transferable and can be used in a variety of working environments and situations. Many people with a degree in statistics decide to change career path because it is relatively easy to do this.

9. What do you think are the main challenges in teaching statistics today?

I think one of the biggest issues when teaching Statistics is handling the anxiety that can come with Mathematics. I think it is crucial to always highlight the importance and the beauty of the mathematical concepts and how these concepts are naturally implemented in the statistical and probabilistic framework. Generally, statistics do not require overly sophisticated mathematical concepts. I think that statistics can encourage people to study mathematics as it can provide real-world contexts where mathematical fields like calculus and algebra “come alive”.

10. Who are the people who have been influential in your career?

The people involved in the working group on model based clustering (which is including my PhD supervisor Prof Brendan Murphy, Prof Adrian Raftery and Prof Gilles Celeux) have been very influential in my academic. There are a lot of important things I’ve learned from the way statistical research was carried out by these. These include the importance of sharing ideas, listening, teamwork and collaboration.


Copyright: Image appears courtesy of Professor Gollini