New professor Elise Dusseldorp: ‘The longer you’re in research, the more humble you become’
Elise Dusseldorp has been appointed Professor in the Methodology and Statistics of Psychological Research. In the same way that she spends her spare time rambling through the forest, as a professor she sifts through colleagues’ research data. ‘I often come across information that doesn’t appear in the manuscript.'
You might ask what trees have to do with statistics? Dusseldorp’s answer would be: ‘More than you might think. There’s a particular statistical method, recursive partitioning, where you build a tree model. You start with the trunk, which represents the average in the data. Then you split that up: you look, for example, at which patients respond well to a treatment, and which ones definitely don’t. You then divide these groups further until you eventually end up with a tree shape: the tree-based model.’
Trees have a dual meaning for her. ‘I love nature. While other people have something with water or flowers, I have that with trees. They calm me down; they emanate something special because they are so old. It may sound a bit airy fairy, but I believe they know a lot.’
Dusseldorp also draws wisdom from the ideas of researchers who preceded her. There are four quotes on her website: what do they mean to her?
‘The best thing about being a statistician is that you get to play in everyone’s back yard’ - John Tukey
‘I think it’s not just fun working with data, but also getting to know the story behind a research project. Researchers sometimes say they have identified an effect, but then when you look into their data and analyses and get to understand their approach, you see that they have also identified other, different effects - information that didn’t appear in their narrative. I think it’s important to see that background: does the story on paper do justice to all the data and the analyses?’
You work mainly in the field of health psychology. How did you get into that?
‘When I did my Psychology master’s here in statistics, we studied data about dolphins or flowers. “Where is the patient? Where are the applications?” I asked myself. I was offered a student assistant position in Health Psychology. During that time, I really enjoyed working with Stan Maes, Winnie Gebhardt and Jacqueline Guicherit, who now works in Clinical Psychology. As a statistician, I was involved in all research studies. I was very keen to stay in this field, so Stan encouraged me to do a PhD.
‘Since then, I’ve published a lot on health and clinical psychology, where I combine algorithms with statistical knowledge: on the effectiveness of therapy, for instance. We often know for a particular treatment whether it works well on average, but you want to find out for which specific groups it works best for and for which groups less so. In a study on depression, for instance, we discovered that mindfulness is most effective in participants who also suffer from rumination.’
‘Statistics is both the science of uncertainty and the technology of extracting information from data’ – David J. Hand
‘At one time I paid a work visit to David Hand in Imperial College London; he is a very good statistician. What he means by this statement is that you really want to get all the available information out of the data. But there is a contradiction in that: on the one hand, your aim with statistics is to create as much certainty as possible, but on the other hand the margin of uncertainty in information also becomes apparent. When a treatment is effective, but it was only tested in twenty people, you can use statistics to show that there is a lot of spread in the results.
‘The longer you’re involved in research, the humbler you become. I’m so often struck by the limitations of data and research findings. You can falsify data, or show that something doesn’t work, but proving that it does work is a lot more difficult. It may work for the target group you’ve studied, but possibly not for someone in Africa. That’s why I believe it’s important for people to collaborate more and repeat the same study in different areas of the world. That might not be as exciting as always studying something new, but it does advance our knowledge.’
You use increasingly smart algorithms in your research. Do you think it will ever be possible to eliminate uncertainty altogether?
‘No, we always have to keep zooming out to see if the algorithm is still correct and keep on using our own thought processes. When I worked at TNO Research, we developed an algorithm that could use data from baby clinics to predict whether a two-year-old child would end up in special education. But you really can’t tell the parents on the basis of that data that their child will go to a special school.
'Even with algorithms, we have to keep using our own thought processes'
I presented the results on a kind of thermometer with different colours, so the degree of uncertainty was clear. The doctor was able to use this thermometer in the discussion with the parents and tell them that their child might benefit from extra guidance. Closer investigation revealed that the parents had four children and didn’t take much notice of the youngest. Once this child was paid more attention, the developmental issues often disappeared. Algorithms mainly serve as an alarm bell. You can never use an app, for example, to check whether your child will end up in special education; that’s nonsense.’
‘Gathering knowledge to then use it for the general good has to be the primary aim’ - Aletta Jacobs
‘For me, that means that I always bear in mind how useful my research will be for society. Not that I want to develop something that will be put to immediate use at the baker’s; statistics are always more of a background factor. With my PhD thesis, for example, I showed how important it was to provide psychological support to patients who had suffered a heart attack. That first publication is still my most cited paper. Because these kinds of papers received a lot of attention in a short period of time, the importance of psychological interventions filtered through to the medical world a lot more rapidly.
‘At TNO we also did research on the building characteristics of psychiatric institutions. It was discovered that the design of the institutions affected how frequently patients were placed in seclusion. Limited visibility for nursing staff into the communal areas and a lack of places where patients could retreat resulted in more admissions. With that research we were able to recommend that greater care should be taken in how these buildings are designed.’
Are your recommendations then immediately put into practice?
‘No, that’s the hard thing. I can’t go knocking on the door of the Minister and asking, ‘Have you read this article?’ But this kind of publication does raise awareness of issues, and in most cases change comes about indirectly.’
‘To choose time is to save time’- Francis Bacon
‘This is a very practical one for me. I have a lot of enthusiasm, so I’m often working on too many projects at the same time. But there’s no point in doing five things simultaneously. And not only that, I realise that when I concentrate fully on something, I’m a lot quicker. These days, I mark out my time more consciously, including time when I don't have appointments and I seek quiet and calm. Having said that, though, I see around me that there is little room for that.’
'I see around me that there is little room to seek quiet and calm'
How do you see that?
‘With students, I can see that they are being lived. And that includes my own son, who’s almost 22: he studies, has a part-time job, is active in a study association, does rowing in between and goes out to the bar in the evening. While I was doing my PhD, I also worked through the weekends; everything had to serve a purpose. I dreamed of “just” becoming a saxophonist in a big band in my free time. When around that time I started hyperventilating, I realised that you also have to make space for your emotions. I had to learn how to do that. Now, as a professor, I work four days a week, and one day I visit my parents who are in a nursing home. For me, that feels good on an emotional level.
‘This quote reminds me that as humans we have limitations and we have to choose the things that are important for us. The odd thing is, to do that you need to have one day a week doing nothing: allocate time when you don’t know what’s going to happen, when you look for a bit of uncertainty.’