Vanda Inacio De Carvalho
Laura O'Sullivan and Becky Nisbet worked together to produce this article as part of our series of Academic Interviews; featuring Vanda Inacio De Carvalho!
Can you tell me about your early life and growing up?
I was born and grew up a few kilometres away from Lisbon in Portugal. During high school I wasn’t really thinking of maths, I liked the sciences in general, but then during my higher courses I got bored. I was fed up learning about them and didn’t want to spend the next four or five years studying them at University. I’d enjoyed maths, as I didn’t need to memorise things, so I thought I’d give it a try! When you apply in Portugal, you complete a form with six or seven preferences for degrees and Universities. I wanted to go to a specific University so I put all seven degrees for the same one and put maths at the top. I got in and I did my undergraduate studies in Applied Mathematics!
How was the experience?
First year was a transition. I was used to high school; University was different and things got a bit dry. First year was mainly linear algebra and calculus, but second year was more analysis and pure algebra. I wasn’t sure that this was what I wanted to study, but there was no point in changing degrees so I stayed. In third year I could pick education or applied maths. I did think that I’d like to do education, but there weren’t stable career opportunities for high school teaching.
I looked into actuarial sciences and more applied subjects, but it wasn’t where I fell in love with statistics. I actually ended up doing more courses on actuarial sciences and operational research than statistics.
In 4th year I took econometrics and enjoyed it. I then thought that maybe doing an MSc in Economics would not be a bad idea and I went for it. Unfortunately, since I didn’t have prior exposure to economics, courses such as public or industrial economics were tough for me and I ended up only staying for a couple of months on that MSc programme! I went back to my previous University department and did more maths, getting an MSc in Applied Mathematics.
Later and because I also liked biology so I thought I’d try biostatistics, I got in touch with a University of Lisbon professor who did biostatistics and she encouraged me to apply for a four year PhD grant. I was awarded the grant and after four years I received a PhD in Statistics. After it I was collaborating with a colleague in Chile, and a job as Assistant Professor came up there, so thought why not? I went to Santiago in 2012 and stayed there for four years!
I enjoyed that position, but I was missing home and I wanted to come back to Europe. I remember that I saw an advertisement for a job in Edinburgh, so I applied, got it and came here.
What are you currently working on?
In terms of research, my PhD was on statistical evaluations of medical tests. If a Covid test is only 70% specific or sensitive, what does this mean? Are there any external factors affecting the performance of the test? For example, glucose levels tend to be higher in older individuals without meaning they’re diabetic, so you have to take the age of the subject into account otherwise you’ll come to the wrong conclusion and think someone is diabetic when they’re not.
I look at the models behind these decisions. Popular models in the medical literature assume that the probability distributions of the test outcomes in the diseased and nondiseased populations follow a normal distribution. If they do not, then the conclusions that we reach about the accuracy of the test may be wrong. I then work on the development of flexible statistical models which don’t rely on a specific parametric distribution for the test results and also that do not assume a pre-specified form for the way external factors (e.g. age in the case of the glucose levels example) affect the outcomes of the test.
We are now in the era of big data! However, in medicine, and particularly in diagnostic studies, we often have case-control studies, so we recruit individuals with the disease and control subjects. This means that, in general, I don’t have very big samples.
What are your plans for the future?
I would like to keep expanding my research as I’m working in a bit of a niche. I’m still trying to find the right balance between methodological and applied projects.
Every three or four years it’s good to teach something new. I learn a lot when preparing a course and I also improve as a researcher.
Do you have any tips for students?
Find out what you really like because a PhD is around four years. For instance, look at some review articles to get a general idea, not too technical, of some research areas and where they are heading to.
You don’t need to be a top student to get into a PhD. Motivation and curiosity are important. There’s space for everybody, so just find out what you like and go and approach people. Nowadays there are many online resources, which even include research talks, so do some pre-research on what you’re interested in and who you’d like to work with.