‘The world is the statistician’s oyster’: Professor Valerie Isham looks back on her RSS Presidency

31st December 2012 marked the last day of Professor Valerie Isham’s presidency of the Royal Statistical Society before the baton was handed over to Mr John Pullinger, with whom an interview will follow later this month. Here Statistics Views talks to Professor Isham about her two years as President, how the Society has evolved over the years, the International Year of Statistics and the influence of statistics on her own career.

(Editor’s Note: Professor Isham is in fact the 100th President of the Royal Statistical Society as some of her predecessors have served twice.)

With an educational background in statistics with a PhD from Imperial College and you are currently Professor of Probability and Statistics at UCL where you have taught since 1978, when and how did you first become aware of statistics as a discipline and what was it that inspired you to pursue a career in the field?

I studied mathematics at Imperial, and at that time there were not many modules in statistics. So I did a couple of courses as an undergraduate and one of them focussed on stochastic processes, and that attracted my attention – so not so much statistics per se but probability and stochastic processes. I’ve always really enjoyed working in that area; I like the challenges of developing mathematics for its own sake but also being able to apply the understanding to all sorts of physical processes.

What I have tended to do in part of my work is to develop models or processes that may be useful building blocks for applications. One of the things I did very early on with colleagues was a model for a particular point process where there is a regulatory mechanism built in. This was a neat process which we thought at the time would be fun to explore but actually has turned out to be widely applicable as a model for earthquake occurrences. Other early work is now cited in modern molecular evolution and estimating species divergence.

You can play with mathematical models and come up with things that are interesting, but I also love the kick of having done something that later turns out to be useful. With many of the things I do, I start off working on a theoretical problem and then work with collaborators in applied fields to answer practical questions.

What do you think the future of teaching statistics will be? What do you think will be the upcoming challenges in engaging students?

I suspect that for really good students wanting to learn about mathematics and statistics in depth, one is not going to make significant changes. There will always be a role for face-to-face teaching and lectures. However, there will also be a large role for online teaching. When you think about teaching statistics to students for whom it is a tool but not their primary interest, you’ve got to attract their attention, motivate them and make them see the power of the subject so that they want to learn. There I suspect there will be a need to use more interactive computational tools. Multimedia presentations are now all-pervasive, and things that would have been impressive twenty years ago are not now. This changes people’s expectations and has an impact on education.

Students will not want just to see a simple graph; they will want the kind of visual statistics that Hans Rösling presented in ‘The Joy of Stats’. On the one hand this is exciting but the lecturer will require a lot of tools and time to put these audio-visual presentations together. Nevertheless, there will still be a role for pencil and paper. Ultimately mathematics and statistics are not spectator sports; you have to be able to do it yourself and not just watch someone else. A lecturer needs to take time to work examples through slowly and carefully, and explain the thought processes that are involved. How do you get started? – that is what students find hardest. I think that’s the part that is difficult to do any other way – it is not just a matter of watching but of really engaging the brain.

Your current research focuses on applied probability – the development and application of stochastic models. What are your main objectives and what do you hope to achieve through your research?

The objective is often to use mathematics to gain understanding of a physical process. This might be the spread of an epidemic e.g. HIV, or the link between rainfall, soil moisture and biomass. If you really understand the process, then you have a good chance of controlling it, and whether that’s flooding or the spread of an infection, ultimately if you can have an effect on the real world, that is very satisfying. I have had the opportunity to work with some wonderful collaborators. It is really good to work with people in applied fields who have a real interest in the results obtained. In a lot of the work that I have done, the assumptions have been motivated by mathematical tractability but it is doubly satisfying when you can answer a real problem and the results have a practical importance.

You have authored many publications. Is there a particular article or book that you are most proud of?

I enjoyed writing Point Processes with David Cox which was published in 1980 quite soon after I completed my PhD (on spatial point processes). There was a big explosion of interest in point processes in the 1970s and the subject was expanding fast. I really valued the opportunity to work with David on the book and learned a huge amount.

Do you think that mathematics and statistics undergraduates and postgraduates starting out today are under more pressure to publish and to obtain grants than when you were a student yourself?

There is a huge pressure to publish but it’s not on undergraduates or postgraduates. The pressures on them are different – to do with funding, employment opportunities etc. It’s not until postdoctoral levels and beyond that there is a pressure to publish and the need to obtain grants. Mathematics and statistics are not expensive subjects as there is no need for large-scale equipment. Thus there has not been the same need to obtain grants as is traditional in many other subjects. Most of us still do not work in teams and our work can usually be done with pencil and paper or on a computer. Of course universities need to bring in the money to support all research and there is pressure to obtain grants to support postdoctoral researchers and funding for research students.

Since the advent of the RAE (Research Assessment Exercise), which has now become the REF (Research Excellence Framework), there has been a lot more pressure on academics to publish. Traditionally people would tend to write a relatively small number of important, substantial works (this is especially true, for example, for pure mathematics). These assessment exercises tend to result in pressure to publish more frequently, which can work against scholarship. You write something that you can publish relatively quickly with, if appropriate, a follow up paper devoted to further developments, rather than waiting and writing one more substantial and complete account. There is also a pressure to work on things where a publishable outcome is reasonably certain rather than on more speculative topics.

Do you have any advice for students considering a university degree in statistics?

I won’t be popular with some of my statistical colleagues when I say this but I would say study mathematics first, although it does depend on what you want to do. If you want to contribute to the methodology of statistics, you really need a reasonable level of mathematics, which you have to acquire somewhere. The easiest way to do this is to study mathematics as an undergraduate. Even if you are not interested in the methodology and are more interested in the applications instead, the firmer and more secure the theoretical base that you have, the easier it will be to work in applications. You need to be able to read research papers, some of which are highly mathematical. Many statisticians do come through other subjects, such as biology or medicine, and work first in the applied area and then learn the theory afterwards but that is harder to do. So my recommendation is to get the hard work done first!

Over the years, how has your teaching, consulting, and research motivated and influenced each other? Do you continue to get research ideas from mathematics and statistics and incorporate your ideas into your teaching? Where do you get inspiration for your research projects and books?

My research projects follow mostly from other research – journals, seminars and fellow collaborators – thinking about “what if” questions, and from working with research students. Some of my research does sometimes spill down into teaching, e.g. I have run a course for PhD students through the London Taught Course Centre and some research ideas go in at that level. My research does not really feed into my undergraduate or Master’s teaching but of course I may use some ideas from real world applications that I have been involved in to make it more interesting for the student.

What has been the most exciting development that you have worked on in statistics during your career?

I don’t think I can pick one! For me, it is this opportunity to see something I have worked on earlier becoming important and widely used. For example, the rainfall work we did in the 80s is in widespread implementation around the world. I wouldn’t call that work mathematically deep, the models have a simple tractable structure and the properties can be obtained relatively easily, but it has had a great impact which is very satisfying.

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?

Over the last two decades, computational advances have revolutionised the subject. In the past, one was very dependent on what one could do algebraically and analytically, which is still hugely important in gaining an understanding of processes. Nevertheless, if you want to analyse large data sets, it is only now that you have the opportunity to do so. We can now begin to make sense of the complex structure in these big data sets, and therefore to make important advances in research. In areas such as environmental sciences and medicine in particular, huge amounts of data are being collected and we are now in a position to analyse them. This generates opportunities for developing both new theory and methodology, and also the means to implement the methodology.

For the future, genetics and genomics are hot topics and lots of interesting work is being done in financial security and on networks (for example networks are involved in the spread of human, animal and computer viruses, in the spread of information and viral marketing, in cell regulation, in systems biology and in social networking). There are many theoretical challenges but lots of subject matter advances to be made. I think the world is the statistician’s oyster at the moment!

What do you see as the greatest challenges facing the profession of statistics in the coming years?

I think the challenge is to decide what statistics is. There is no clear boundary to statistics as a subject – it overlaps especially with economics, computer science and the social sciences. In some cases, the wheel is being reinvented. But I think the main challenge is to improve the image that statistics has in the world. I think people don’t fully understand the strength and capability of statistics and think it’s nothing more than adding up numbers and the simple presentations of data. We need to ensure that the best use is made of statistics and to maintain the subject as a discipline, not to repel encroachment from other fields but to work with them to make advances. Statistics is probably the most interdisciplinary subject there is.

Are there people or events that have been influential in your career?

First and foremost, as mentioned earlier, David Cox, who was my PhD supervisor and with whom I have worked on many projects over the years. Then there is Ignacio Rodriguez-Iturbe, who is a very distinguished hydrologist at Princeton – a fantastic person, a great enthusiast and real inspiration to work with. Also now at Princeton, but formerly at Cambridge UK, Bryan Grenfell, who is an ecologist, with whom I worked on the transmission of macroparasite infections. Working with each of them is hugely enjoyable – I have learnt a lot from them and am extremely fortunate to have them as collaborators.


Copyright: Photograph appears courtesy of Professor Isham