As a biostatistician, she works in an environment of constant and rapid change. New Research Award winner Rebecka Jörnsten sees challenges with our new big data existence.
Winter has begun to tighten its grip on Gothenburg, and Rebecka Jörnsten’s office at Campus Johanneberg is so cold that she’s wearing not only a heavy cardigan but also gloves indoors. She huddles up, smiles and squirms somewhat self-consciously when she’s about to answer the question.
“It feels really great, of course, to have received the award. I interpret it as an appreciation of the fields of statistics and biostatistics, the importance of which is constantly growing with the challenges of analysing large volumes of data. I’ve worked extensively in these areas.”
IT’S MID-NOVEMBER and the day after Rebecka Jörnsten received the Faculty of Natural Sciences’ Research Award. The explanatory statement indicates that among other things she introduced the big data paradigm in the formerly more traditionally oriented statistics environment in Gothenburg.
Historically, access to data has been a problem for statisticians, but in recent years that has changed radically. Today there are huge amounts of data available, providing entirely new opportunities and challenges for Rebecka Jörnsten and her colleagues.
“With small amounts of data, it can be difficult to draw conclusions that are statistically significant. With large amounts of data, it becomes ‘too easy’ if you don’t think about the big picture. Big data can be difficult to work with since it often is unbalanced and compiled in a motley fashion from different sources.”
ALMOST NINE YEARS have passed since Östergötland native Rebecka left Rutgers University in New Jersey and 14 years as a researcher in the United States to move back to Sweden. She received a job as a researcher at the Department of Mathematical Sciences. The large statistics division at the department attracted her.
“It was important for me to join a team that had the kind of dynamics that exists here, that all subjects are represented. And I liked Gothenburg’s small-town charm. But at first I thought it was strange that people smiled at me so much. I thought, ‘Have I spilled something on myself?’”
IF BIOSTATISTICIAN Rebecka were to describe her research in a word, it would be “model selection”. To be more specific, this involves developing statistical models for large-scale data and how the models should be compared, visualised and interpreted. One of her major collaborations is with researcher Sven Nelander at Uppsala University on the development of network models describing genomic variations among different types of cancer. The goal is to find biomarkers that can predict how individual patients will respond to treatments for various types of cancer. A biomarker can be described as a measurable indicator of a biological condition.
“One of the best things about being a statistician is precisely that it’s possible to find opportunities for collaboration with researchers in many different fields. For example, I’m in a medical collaboration about how music can be used as an alternative form of treatment.”
SHE CURRENTLY is leading a team with two doctoral students, and the group will grow further. She has received research grants that will be used to expand with two doctoral students and a post doc, but says it’s difficult to find good people.
“Trained statisticians are very attractive on the labour market, and when they’ve finished their education, they quickly find employment. At the same time it’s hard for me to envision building up a huge lab. I like the smaller setting and having direct personal conversations.”
While the large amounts of data create new opportunities, developments within academia also have moved towards faster and larger volumes of published results. Which, in turn, makes it more difficult to find the right things, and places higher demands on criticism of one’s sources and critical thinking. She compares this with how things were when she was a doctoral student.
“Then I went physically to the library and picked out a journal with an article that I wanted. Today it goes so terribly fast. Enormous amounts of results are published in a growing body of scholarly journals, and it becomes very difficult to have an overview of how much there is and what the hidden gems are.”
SHE BELIEVES THAT IT can be hard for researchers to know if they are at the forefront or being left behind, which can be a stressful. The increasing availability of data and results is something Rebecka Jörnsten emphasises for her students, and this leads to new development of statistics courses.
“I always try to make my courses as ‘hands-on’ as possible so that the students get to both learn something and apply it. Then a student can already do a lot with actual data on the first-cycle level and later in courses with challenging analyses on large volumes of data from different fields. But just as in research, you have to look at the available material with a critical eye. Codes and online results are not always a guarantee of quality.”
Family: Daughter Ingeborg, 4.5 years old
Place of residence: Central Gothenburg
Occupation: Professor of Biostatistics and Applied
Statistics at the Department of Mathematical
When I'm most happy: “When I do something
with my mother and Ingeborg. And when we have
brainstorming meetings in my research group. Or
when a student comes up after a lecture saying ‘I
hadn’t thought about it in that way,’ that makes me
Talent: “I'm good at cooking. Not baking since that
requires me to do everything perfectly or the results
are a disaster, but other cooking is fun.”