Christmas is coming, corona infections are increasing rapidly, but politicians and hospitals are not worried, they say: the healthcare system can handle it. How different it was four years ago, when humanity suddenly found itself in a pandemic. Policymakers everywhere wondered: how should citizens behave to keep the misery within limits, and above all: how do you get them to do so?
If you want to provide effective information about this, behavioral scientists quickly said, you must first look at which groups in society trust which (religious or local) leaders. Also remember that standards for good behavior are more effective if people in your own ‘bubble’ adhere to them and if you expect them to appreciate it when you do the same. And don’t call keeping your distance social distancing but physical distancingbecause social contact is still possible.
An international group of 42 behavioral scientists already did this in April 2020 in total nineteen recommendations in the scientific journal Nature Human Behavior: non-medical tips for health experts, policy makers and politicians. They naturally relied on social scientific knowledge that had been collected before the corona outbreak.
Research in a lab
So was what they said correct? Broadly speaking, yes, according to an evaluation by the same scientists, who investigated it together with about forty colleagues. Sixteen of the nineteen recommendations received research support, to conclude them in magazine Naturealthough most effects were not very large.
In addition to the importance of group feeling and trust in leaders, the advice included target group policy, cultural differences and the dangers of polarization and disinformation. Two recommendations actually turned out to have no effect and one (the term physical distancing) no research had been done.
The reason for the evaluation was criticism of the 2020 article. Although it was cited thousands of times by other scientists and used by governments to shape measures. “But there were also people who wondered: is that science robust, are those psychology studies replicable, and can you simply translate research done in a lab into the field,” says Naomi Ellemers (Utrecht University) on the phone . She is one of two Dutch scientists who co-authored both articles; Paul van Lange (Vrije Universiteit Amsterdam) is the other.
Ellemers partly understands that criticism. “Although people sometimes seem to forget that medical science also had to extrapolate from other knowledge to corona, because corona was unknown at the time. Anyway, two years later there was all this new research and our 2020 recommendations had essentially become ‘pre-registered hypotheses’.” Recording your hypotheses in advance is a condition that is now often imposed on medical and behavioral science research to prevent researchers from presenting the effects found afterwards as if they had been predicted..
The evaluation process followed a complicated method. They had each claim investigated by two teams. Team one consisted of authors of the first article (who may have had an interest in ensuring that their previous claims were correct), while team two consisted of a new group of also about forty independent scientists. Everyone was assigned a claim and had to say, based on new corona-related research (747 articles in total): is there evidence for it, what kind of evidence and how strong is it? The teams agreed with each other for 77.5 percent: a high percentage for this type of evaluation, according to the researchers.
As mentioned, support was found for sixteen of the nineteen recommendations, but usually the kind of support for which the behavioral sciences are criticized: intentions are often examined (‘what would you do if?’) instead of actual behavior, effects are often small and context-specific, and there were many questionnaire studies and few experiments, and even fewer ‘in the wild’, among people outside the lab.
Group norms shape behavior
“We indeed need to do more research with creative, concrete behavioral measurements,” says Paul van Lange. He audibly enjoys linking the claims to what he saw in practice during corona, for example about how group norms help shape behavior, the claim with the most support in the article: “During corona you couldn’t rely on your habits, so you focused on what you saw others in your own group doing. This partly explains why young people quickly developed different norms and behavior than older people. I walked a lot in the Vondelpark at the time, and there you saw that young adults were less adhering to the rules.”
Ellemers emphasizes that similar problems exist in medical science as in behavioral science: “There too you often have small effects that add up.” And new medicines are not tested directly in patients, in ‘the real world’, but first in the lab and in healthy volunteers. Ellemers thinks that people are often inclined not to take behavioral sciences so seriously. “That is why I find that analogy with medical science valuable.”
Unlike medical science, it is not a standard practice for behavioral policy to first conduct practical research. There are some initiatives. Van Lange gives workshops to mayors in the Netherlands and is a member of Adapt Academy, a partnership of four universities for knowledge to help society weather crises. Ellemers and her colleagues work with institutes such as the Social and Cultural Planning Office, where people have half a job there and half a job at a university.
This prevents, which according to Ellemers often happens, someone saying afterwards: oh yes, let’s measure how good this policy measure was. “If you don’t have a preliminary measurement or control group – what? There are examples of very expensive projects, for example to help people find jobs, that are never investigated to see whether they work.”
Reading list