Dr. Martine Barons is a mathematician who works at the Department of Statistics at the University of Warwick, having joined in 2013. Her research interests include mathematical modelling and analysis of human systems, decision support, machine learning, complex systems and data analysis. Martine spent the Autumn 2015 term in Melbourne, Australia working with biosecurity and other experts on decision support in food security and pollination.
“…Maths and statistics have a reputation for being male dominated and this certainly matches my experience. At undergraduate level, whilst the student ratio was 50:50, all my lecturers were male. As a postgraduate student, I was one of three female students out of 13 in my cohort, and that was a ‘good’ year…”
A relatively recent career move into maths
My career in mathematics began relatively recently. At school I was told I wouldn’t cope with A-level maths, so I took economics instead and went into accountancy. I gave up paid employment when my husband and I had our children – having myself had both parents work full-time had a lasting, negative affect on me and I didn’t want that for my own children.
When they were at school, I did a mixture of paid and voluntary work which allowed me to be home in school holidays, look after them myself when they were ill and be free to do the school run at both ends of the day. As you might guess, the work available was low-paid and insecure; it shouldn’t have to be that way.
As our youngest entered teenage I realised I would need a pension and therefore some steady paid employment. With governorship at two schools and voluntary work as a schools speaker for an AIDS charity under my belt, I decided to become a school teacher.
Maths can make people’s lives better! Who knew?
I enrolled at my local college to take A-level maths with some trepidation, but worked hard and did well. I had offers from both Warwick and Coventry universities and picked Coventry because it had earned a centre for excellence in teaching and learning of maths award.
It was there that I was introduced to mathematical modelling and experienced a paradigm-shift: I discovered that maths can be used to make people’s lives better! Who knew? I was astounded! It was very early in my first year of undergraduate that we investigated the claim being made by some newspapers that there was about to be a second wave of cases of vCJD [Variant Creutzfeldt-Jakob disease] – mad cow disease. Our simple mathematical model showed there wouldn’t be a second wave – and there wasn’t!
A three-year degree – not enough time to study so many wonderful topics
It soon became clear to me that a three-year degree was not going to be enough to study all the wonderful topics that exist and one of my lecturers suggested I look at jobs.ac.uk. It was there that I discovered the studentship – a mechanism that allows for a student to be paid a stipend whilst they study for a higher degree.
As soon as I saw the studentships advertised at the Warwick Centre for Complexity Science, I knew that was what I wanted to do. I doubted they would consider me, and I nearly didn’t apply, but decided I needed to set my youngsters a good example, so I went for it. I was surprised enough to be called for interview and went into shock when I was offered a place!
One of my undergraduate lecturers felt it only fair to warn me that I wouldn’t cope with this level of maths, but I decided to give it a go anyway. It was certainly very hard work at the Warwick Centre for Complexity Science, but I passed my Masters, then earned my Ph.D. and was taken on as a research fellow in Warwick’s Department of Statistics.
My role at the University of Warwick on a day to day basis
In academia, to pursue a research idea you apply for grant. If the funding body thinks it worthwhile and it is funded, the applicant, called the principal investigator (PI), can either reduce their teaching and admin roles to do the research themselves or employ a research fellow (RF) to work on it full-time whilst they advise.
I am employed as an RF on an EPSRC [Engineering and Physical Sciences Research Council] grant to work out how to network together information from disparate panels of experts in order to provide decision support.
For example, to make decisions about policies to reduce food poverty it is necessary to combine information about weather and climate, crop yield, international trade, population demographics, consumer behaviour and economics, to name but a few.
An expert panel on each of these can make forecasts about their particular domain of expertise, with a degree uncertainty, but it is not straightforward to combine these together coherently and score different policy options, taking full account of the uncertainties. This is what I am working on with my PI.
Synthesising research and thinking hard about the problem in hand
On a day-to-day basis, my role is quite varied. Obviously, the bulk of my responsibility is to read and understand what other researchers have done and to think hard about the problem in hand, trying various approaches, evaluating their usefulness and writing research papers.
It is also important for scientists to talk to each other about how we are approaching a variety of problems to share insights that would help each other. These workshops and conferences need to be organised and that is another thing I do – at its core this is not so different from organising children’s birthday parties except that the activities on the day are different.
There are also community-type activities to be done, like helping at open days, advising students, talking to the press, reviewing other researchers’ papers, collaborating with external partners and writing applications for research grants.
Putting theoretical decision support work into practice in Melbourne
I recently went to Melbourne to work on a project relating to bees. This was to start putting the theoretical decision support work into practice, showing it really has something to offer on a policy matter that is highly pertinent at present.
As you may know, there is concern about reductions in the number of honey bees, wild bees and other pollinators and the effects this may have of the security of global food supply. DEFRA [the Department for the Environment, Food and Rural Affairs] produced the National Pollinator Strategy in late 2014 to protect both biodiversity and eco-system services such as pollination in the UK. I am in the process of building a decision support system to score different policy options to aid decision-makers in selecting the best policy for their priorities.
One of the difficulties is that insect pollinators have not been well studied in the past with respect to their abundance and responses to changing environments. My plan in Melbourne was to work with experts in population distribution modelling. These mathematicians and statisticians work on how to infer the size of the underlying populations based on various types of surveys.
If you think about it, if I see a bumble bee in my garden and record it, that doesn’t tell us much about how many bumble bees there are in my garden, except that there is at least one. This kind of presence-only data is particularly problematic, but a large body of the data available is of this type.
Unexpected opportunity: Research doesn’t always go the way you plan
However, research doesn’t always go the way you plan, and in my first week in Melbourne I met a highly knowledgeable and well-connected purveyor of beekeeping supplies. He was very excited by what I am working on and made use of his extensive network of contacts to furnish me with introductions to experts in government policy, commercial beekeepers, bee breeders and researchers.
I capitalised on this unexpected opportunity and spent a lot of time interviewing these experts about the pollinator problem, the eco-system and its effects on pollinator populations and various government policies and their effects. These insights were extremely useful in pushing the development of the decision support system forward. Since returning to the UK, I have organised a workshop for UK-based pollination experts to elicit different aspects of the problem.
Other practical applications for my mathematical and statistical knowledge
I have applied my mathematical and statistical knowledge in a variety of other problems before I began working in decision support.
I worked on traffic models, investigating those irritating stop-and go waves, also called ‘phantom jams’, when the traffic comes to a halt and you imagine there is a lane closure or something up ahead and it turns out there is nothing. I used data from the M42 to ensure my model was producing realistic predictions.
I worked on cardiac rehabilitation data and published my research in the British Medical Journal’s BMJ Open, showing that when you measure fitness and take that into account, BMI tells you nothing about how long a person using cardiac rehabilitation services will live.
I also recently provided statistical support for an investigation on how flexible construction material, used for bridges etc., affects human balance.
Breaking down barriers for women in maths and statistics
Maths and statistics have a reputation for being male dominated and this certainly matches my experience. At undergraduate level, whilst the student ratio was 50:50, all my lecturers were male. As a postgraduate student, I was one of three female students out of 13 in my cohort, and that was a ‘good’ year – one in ten was typical and this is the proportion I see in maths academic staff. In the Statistics department, almost a third of the RFs [research fellows] and almost a quarter of the permanent academic staff are female.
Having entered academia with a wealth of prior experience in other places of employment, the last thing I was expecting was to find that being female would mean I would sometimes be treated differently than my male colleagues. What I have seen and experienced has been very subtle and I thought at first I must be imagining it.When I decided to deliberately observe carefully, I noticed that, for example, female speakers were more likely to be interrupted by members of the audience than male speakers, who were more likely to be allowed to finish their talk before taking questions.
The nature of the questions also seem to vary, for example, when asked about the relationship between their work and other researchers’ work, typically I observed a male speaker asked at the end of his talk something like, “Smith & Jones published a paper on this topic, how does your work relate to theirs?” whilst often I observed a female speaker interrupted mid-sentence with something like, “but Smith & Jones have already done this…”. This kind of treatment is something I have sometimes experienced myself, but not by any means every time I give a talk.
Help from Athena SWAN
I think the adoption by the funding bodies of Athena SWAN awards is beginning to tackle these last vestiges of sexism in the mathematical sciences. These days, departments have to have at least a Bronze award to receive grant funding and this has meant examining how things are done and whether it might make life more difficult for some, for example, those with caring responsibilities.
Often, it is simply a matter of not timing meetings to clash with the school run or giving plenty of notice for evening and weekend events. Each instance of making it more difficult for some is small by itself, but the cumulative effect can be quite significant and even cause those affected to drop out of academia. The fact that I waited until my children were grown up before embarking on a career in the mathematical sciences has made me immune to most of these organisational-level problems.
Advice from other women in maths
I can’t think of a specific piece of advice I have been given as a woman in mathematics that has, by itself, made a big difference, but I have certainly benefited from various Women in Mathematics or Women in Science events run by the University of Warwick, the London Mathematical Society (LMS) and the Institute of Mathematics and its Applications (IMA).
It is useful and interesting to hear about the experiences of other women in our male-dominated niche, the obstacles they have faced and how they have been overcome. That said, it is always enjoyable to spend time with other mathematicians, male or female, as what we have in common is far greater than any differences that may exist.
Careers in maths and statistics: From biology to finance, to computing and big data
The best way I can think of for girls and women to find out more about a career in maths and statistics is to come along to popular level talks, which give an idea of the wide variety of different ways maths is used, from biology and finance, to computing and big data. My advice for boys and men wanting to find out about a career in maths and statistics would be identical.
The Big Bang Fair is great fun and shows a variety of careers within the mathematical sciences as well as offering some brilliant activities to try. The LMS runs popular talks and the IMA has local branches which organise popular-level talks and other events across the country. The IMA also has an early career section, as does the Royal Statistical Society and some IMA branch members are willing to offer mentoring of young mathematicians if requested, so there is support available.
There are also some universities that run popular maths and statistics talks, the University of Warwick and Birmingham University to name but two.
Decision support systems for large, multi-faceted complex systems – still a lot to do
There is still a lot to do to fully develop these types of decision support systems for large, multi-faceted complex systems. Developing the new methodology required and operationalising them in collaboration with end-users will keep me busy for a good while yet, I expect.