Subscribe Bookmark



Aug 5, 2014

Celebrating statistics with Denise Lievesley

In celebration of the International Year of Statistics, we reached out to a few British statisticians to ask them to share a few thoughts on statistics as a discipline, statisticians and applied statistics. The United Kingdom has a rich statistical history, and we are pleased to share some of these most interesting blog interviews.

Our third interview is with Denise Lievesley, Head of School, School of Social Science and Public Policy, King's College London. Lievesley, one of the United Kingdom's leading social statisticians, is dedicated to campaigning for the evidence-based development of sound public policies in the UK and beyond. Her distinguished career includes posts as founding Chief Executive of the English Health and Social Care Information Centre; Director of Statistics at UNESCO, where she established its new Institute for Statistics; Director of the UK Data Archive; and professor of research methods in the mathematics department at the University of Essex. She was also the President of the Royal Statistical Society (1999-2001) and of the International Statistical Institute (2007-09). She will be a panelist at Discovery Summit 2013 to discuss the International Year of Statistics.

Question:  How did you become interested in statistics?

Denise: I studied statistics as an undergraduate because I was very good at, and a bit nerdy about, maths. At the time, I didn’t know what statistics was, but I was trying to find a subject that would use a lot of maths and could be applied with a particular relevance to society -- although I wasn’t sure I wanted to be a teacher.

My father was in the Armed Services, so I had an opportunity to travel around quite a bit. Because I didn’t stay at any one school for very long, I was able to choose A levels for myself, so I did all the maths I could. In addition, I picked other subjects I enjoyed, like English (because I love reading) and art (because I love drawing and painting). I know that’s a peculiar combination of interests. But it turned out to be perfect. Maths is obviously useful to statistics, but English helps with communication. And, of course statistics is about patterns, providing a connection to art.

When I was interviewed for a university place, I was told that normally candidates would have studied maths, physics and chemistry, and that accepting me into the statistics curriculum was a bit of a risk!

Question: Can you comment on the relationship between mathematics and statistics?

Denise: Mathematics is of course fundamental to statistics, but there are differences in the philosophies of the two fields. Statistics is the science of uncertainty, and mathematicians aren’t often comfortable with uncertainty. For applied statisticians, it’s not necessary to build the competence to develop theory, but nevertheless one does need to have a broad understanding of the underpinning mathematical theory in order to ensure sound science. Of course the diversity of statistical applications means that there are many branches of the subject. I consider myself to be an applied social statistician, and I love the fact that I can connect with researchers working in so many different fields.

Question: Do you have any recommendations on bridging the communication gap between statisticians and other stakeholders?

Denise: Well, unfortunately many statisticians don’t communicate as well as they should. Take statistical output: Even a simple table of numbers can be interpreted so that someone who is perhaps number-shy or extremely busy can take a clear message from the table and can draw conclusions about what actions might be needed. Most importantly, statisticians should provide the story of the data. But unfortunately, we often start out talking about data limitations, even focusing on why the data that’s been collected can’t be used instead of seeking ways to maximize its value. And often, too much detail is presented, meaning that users cannot extract the key messages from data. Thus, we have lots of work to do to improve our communication skills.

We also need better training for our students and for junior official statisticians as they start working in government. They need to be able to write briefing papers for senior public servants and ministers who decide policy. Some briefings are filled with extraneous detail, or use impenetrable language, or concentrate only on rather esoteric caveats. This reminds me of a T-shirt from the ASA: “Being a statistician means never having to say you are certain.”

Question: You are well-respected as one of the UK’s greatest social scientists. You’re also the first woman to serve as President of the International Statistical Institute. What would say to encourage more women to choose a career in statistics, traditionally a male-oriented field?

Denise: I am not sure that it is a predominantly male field anymore. I was asked to speak at the  Young Statisticians Conference to talk about my career, and I was very pleased to see the large proportion of women present. More and more women are studying statistics, though there are still some gender divisions by application and field. So young women are more likely to choose careers in medical and social statistics, rather than the more theoretical and mathematical statistics paths.

There remains a problem in some countries whereby girls are less likely to take mathematics and thus are less likely to have exposure to statistics. We don’t understand why girls shy away from mathematics, because there’s absolutely no evidence that they are weaker in the subject. When girls are taught separately from boys – as is often done in Catholic countries – girls are more likely to take mathematics and science. If girls are taught with boys, they seem to define these subjects as the male terrain!

I would say to young women: “Be confident.” As it now stands, they seem to feel they have to excel in all areas – socially, as homemakers, as wives and mothers, in their careers. In addition to this, they worry about their appearance, and so they push themselves really hard. It really concerns me. I would say to them that it’s OK to take a shortcut. You don’t have to be fantastic at every area of your life, and you don’t have to do everything so brilliantly.

Secondly, I would say that statistics is a fantastic career. You can specialize, or you can do what I’ve done – every few years I changed direction in a new field of application. The analytical skills you build up are really critical to a wide range of work activities. You can use the skills you have as a statistician every day in terms of analyzing financial data, examining KPIs or working out how to do performance reviews.

Question: Can you comment on the strong connection between the statistical concepts you cover and how you make them relevant for real-world problems?

Denise: At UNESCO, we had responsibility for collecting global data on educational policy and were responsible for reporting on a number of the millennium development goals. In this role, we  called countries to account. This fed into a "democratic" process enabling the public to view the progress of their society. There are many challenges in achieving comparability spanning over both space and time. You have to work at a country level, yet many countries have no functioning statistical system. The job is not just a practical one relating to data collection, but methodological issues around estimation, imputation, data augmentation and synthetic estimation also have to be addressed. For example, how do you measure annual illiteracy rates in every country of the world? It’s just not possible to supply good data annually – it would be far too time-consuming and expensive. If, however, you’ve got a 10-year measure (say from a survey), how do you link it with other educational system data to see if you’re on target for annual change from 2004 to 2015?

We also sought out to measure the number of children in primary school in each country. So, what is the definition of children in primary school? What is the definition of "in school"? Indeed what is the definition of "school"? If half a child’s time is spent taking animals to the water hole or they’re working all day in manufacturing with just two hours of education in the evening, how does one account for that?

So the challenges of the job ranged from raising statistical capacity (through advocacy for statistics) and addressing the practicalities of data collection, through to the diplomacy needed to challenge governments who sometimes lie with their data. The data could be wrong because they don’t have the resources to collect it, they haven’t prioritized statistical work, or it’s uncomfortable for them to say their children aren’t in school. You have to understand enough about diplomacy with the UN system and how you can appropriately challenge and how you can fight for the truth.

And, development is not monotonically increasing. Countries have major setbacks – i.e., major wars and environmental disasters. Who defines progress, and how do we make judgments about what is acceptable?

Sixty million children don’t have access to primary education –  this figure has plateaued in part because in Africa the number of children has increased. The birth rate is very high, and fortunately, we’ve managed to impact early childhood death, so more are living. It may not look like you’ve made progress, but maybe you have. You have to look at the whole picture in context. If you focus on only one element, then you can introduce (dis)incentives inadvertently. There are, of course, already incentives to show a particular picture. For example, perhaps some funding agencies will not continue to provide support to a country unless a certain level of progress has been demonstrated.  In such a case, the country will try to ensure the data show progress. We think purity of statistics is most important thing in the world, but guess what? There are many other factors to consider.

Question: What was your favorite thing about being president of the Royal Statistical Society?

Denise: RSS is really great institution, and it was a privilege to lead it. You get to see the diversity of statistics and get more involved with the whole of statistics profession. I made many right-angle turns in my career, but it’s always been within public sector. So, interacting with statisticians in university and industry was an exciting aspect. The UK is a much smaller country than the US, and this means the RSS is very different from the ASA. The RSS is consulted by the Government (along with sister organizations such as the two Mathematics societies and the Social Science Academy) on many public policy issues. So the RSS has a huge amount of influence, and I really enjoyed that.

Question: Who are your favorite statisticians?

Denise: I really think highly of a lot statisticians. David Cox has to be up there. We’re not in the same field, and we never worked together, but I applaud him for loads of reasons. There was a weekend event where we were both invited to listen to PhD students present on their research work. However weak a presentation was, he always managed to pull out the positives. He engaged with others, energized and encouraged – absolutely fantastic. He would be high on my list. Statisticians who have engaged with others and support younger statisticians – that really matters to me.

I also am in awe of Peter Green, who I worked with when I was secretary of the RSS. I clearly remember sitting in a pub with him once while he explained Gibbs sampling to me. It’s always nice when somebody cares enough to help you understand these things.

Question: What is your biggest professional success of which you are most proud?

Denise: At UNESCO, I was brought in to dissolve a failing division of stats – failing not because of the staff, but because it reported up through the same part of the organization as janitorial, catering and other services. Of course, statistics is a service, but in the same way medical care is a service – a professional service. So I formed a new division, set the work programme and standards, established a governing board, recruited the staff, held an international competition for the country to host us and got it functioning really well. When I left after seven years, some of the ambassadors called the organization “the jewel in the crown of UNESCO.” I don’t take credit – well, partial credit! – but an awful lot of credit goes to 70 bright, committed staff who formed the new organization.

Also, when I was 23 or 24, I developed a system in the UK for using post codes instead of the electoral register, which had many deficiencies. I am so pleased that post codes are still used by market research agencies and government research agencies today.

Question: Do you have any additional words of wisdom that you’d like to share?

Denise: I have a lot of success in my life, but I’ve had messes as well. When I was employed at the National Health Service, I whistle-blew and lost my job. That was a real setback in my career. My email was cut, and my colleagues were told not to contact me. I spent several months not being able to leave my house as I was so demoralised, but then rejoined the UN and went to work in Addis Ababa. At the time, it was a pretty good thing to do because you cannot go there and feel sorry for yourself. It was a really hard place to work, but you can’t think you’re the most important person in the world when you are surrounded with people who have nothing. Still it was very painful, and I was devastated.

My words of wisdom would be that when you see this happening to someone else in the profession don’t abandon them or think that you can become damaged or contaminated by association with them. Also don’t give so much to your work that if some disaster happens you don’t have anything else in your life. You do come out stronger in a sense that such experiences make you realize that it is vital to have a work-life balance.

1 Comment
Community Member

Tom Morse wrote:

Great interview. Her final comments about helping others through hard times is so important at this time with people out of work or underemployed. Thanks for sharing her wisdom and compassion.