Statistics Are Being Abused, but Mathematicians Are Fighting Back

We encounter statistics every day, but it is not always easy to interpret them appropriately. During her doctoral thesis work in the abstract field of number theory, Dutch mathematician Ionica Smeets noticed that it was very difficult for laypeople to find understandable scientific explanations of statistics. So she decided to change that.

Now Smeets authors books, writes for newspapers and speaks about her field on television. She is widely known in her home country of the Netherlands, where she mediates between science and the public as a professor of science communication at Leiden University. There she studies how to present and depict research results so they are as comprehensible as possible. And in her lectures, Smeets teaches laypeople how to debunk misguided statistical conclusions.

Smeets spoke to spectrum of science the German-language edition of Scientific American about misleading statistics, how to avoid them and what it takes to improve communication between scientists and the public.

[An edited transcript of the interview follows.]

What is your favorite example of a misguided statistical conclusion?

In the Netherlands, it was often said that chocolate could cause migraines. I actually know many people with migraines who don’t eat chocolate because of that. A few years ago, however, it was discovered that the mechanism works exactly the other way around: before a [migraine] attack, there are certain reactions in the body that cause you to crave fat and sugar.

When did you realize that you had to fight against such wrong conclusions?

I’ve been giving talks on manipulated statistics for a long time, simply because I think it’s important. The real importance became clear to me most of all when a lawyer approached me a few years ago and told me he had won a case as a result of one of my lectures. I gloated and asked if he could debunk false claims made by his opponent. But he just laughed and said that, on the contrary, he had used what he had learned to create a misleading chart. And he was very proud of it. That’s when I realized: if you show someone how misinformation spreads, you teach them how to do it themselves at the same time.

At first, I wanted to stop giving lectures like that. But then I thought, “You have to talk about these things much more instead.” Because if the other side had known about it, they wouldn’t have been fooled.

Have you ever fallen for a misrepresentation of statistics?

Oh, yes. And I still do. Statistics is an area where you can always get something wrong. It’s so easy to fall for it. If one of my research projects involves a lot of statistics, I always make sure that a [statistics] experience it onboard. Often people assume that if you’re a mathematician, you know statistics—but that’s not true. It’s easy to get confused with probabilities. I’ve learned not to trust my intuition.

What do you have to do to avoid such misunderstandings?

There are studies that have addressed this very question. For example, when a research paper discovers a connection, and the accompanying press release [erroneously describes the connection as implying] cause and effect, this is then usually presented in the same way in the media. If, on the other hand, the university communicates correctly, then, according to studies, most of the media will do the same. That is why it is important to pay attention to correct communication already in the universities. The more precise you become, the better the resulting journalistic articles will be.

I find it very interesting how people blame one another. Universities claim that the media exaggerates issues and doesn’t understand them properly or that it’s the schools’ fault—that children need to be better educated. Journalists, on the other hand, complain that universities are more concerned about their image than their research. Everyone points at the other.

And who do you think is right?

Some things can be improved, but they should start with the universities. Science should feel more responsible. That was the reason why I went back to university. I run a master’s program where students learn how to communicate science well. And we also do research. That’s very exciting, because for many things, we don’t know how they really work—especially if you don’t just want to inform, but you want to get people to change their behavior.

How do you do that?

How you communicate something is extremely important. For example, I was talking to a scientist [colleague] about how anecdotes and stories are much more convincing to most people than numbers. He did not believe me. I showed him statistics and studies about it, but he wasn’t persuaded. And then he actually changed his own mind with an anecdote: As he told me shortly after, he owns a motorboat. My colleague said that everyone knew not to swim in the water with the motor running. In the case of a friend of his, a child got caught in the motor and had to go to the hospital. Fortunately, everything turned out well in the end. But this story made such an impression on my colleague that he was much more careful afterwards. That one anecdote was so much more impressive than all the statistics and rules he had heard before.

How do you scientifically investigate which form of communication is most suitable?

For example, we are looking at the question of how to communicate through videos during a pandemic. The background to this is that the [World Health Organization’s] information videos are very different from the popular clips on YouTube that deal with the coronavirus [that causes COVID]. That’s why we work with filmmakers, technicians and anthropologists, which is very enriching. While scientists focus almost exclusively on the content of a shot, someone else considers, for example, how to get the sound across in the best possible way.

We shot short videos in which the same actor conveys different messages. And then we interviewed test subjects to see how they perceived the messages. We also studied what the perceptions are when the actor appears sometimes as a scientist and sometimes as a salesperson, for example.

Science communication is not always taken seriously—especially if you are aiming for a research career.

There’s even a name for this: the Sagan effect. Carl Sagan was a brilliant astronomer—and yet he was often not taken seriously because he was too popular and did too much on TV. Yet he also did great research and published a lot.

Unfortunately, that is still happening today. I am trying to change this, together with many other people. Not every single scientist is suited to communicate their knowledge to the outside world. But as a department, you should make sure that at least some do.

Are the concerns justified? Do the people involved in communication perhaps lack the time to concentrate on their research?

In fact, there are studies that show the opposite. Researchers who are also involved in science communication also do better in other areas: they publish more; they are cited more often. There used to be a stereotype that science communication was for students who weren’t that good. I was very annoyed by that. Some were even explicitly discouraged from getting involved in the field. Fortunately, that’s changing.

This article originally appeared in Spektrum der Wissenschaft and was reproduced with permission.

We encounter statistics every day, but it is not always easy to interpret them appropriately. During her doctoral thesis work in the abstract field of number theory, Dutch mathematician Ionica Smeets noticed that it was very difficult for laypeople to find understandable scientific explanations of statistics. So she decided to change that. Now Smeets authors…

Leave a Reply

Your email address will not be published. Required fields are marked *