How do our cognitive biases influence us during COVID-19 ?

We do not currently have the necessary hindsight to assess the impact of cognitive biases in the perception and management of the crisis, and it would be very difficult from a methodological point of view to attribute a specific act (for example, not closing schools) to a given bias (such as a bias of optimism on the part of this or that leader). However, it has been shown through many experiments in experimental psychology and neuroscience that certain biases are particularly relevant to explain human behaviour.

1. Exponential growth bias

One of the biases that can have the greatest impact on the management of the health crisis, both at the level of governing bodies and individuals, is the exponential growth bias. Our brain is optimized to add up and process linear growth, not to process exponential growth. However, when a contagious individual infects several people, who in turn can infect several people, the number of cases increases exponentially. This bias leads us to underestimate the speed of spread of the virus when the curve is exponential. Whether the available information is numerical or visual, individuals wrongly perceive the exponential growth of the virus in linear terms.

2. Endogroup or membership bias

This bias characterizes the fact that, in general, members of a group (family, friends) are considered to be more trustworthy, more competent, etc., than those who are not. The closer group members are to each other (e.g., parents, grandparents), the stronger this effect. In the context of COVID-19, there is a misconception that our children are less likely to get COVID-19 and infect us, or infect their grandparents compared with strangers.

3. Optimism bias

Humans tend to be irrationally optimistic, and this has several implications:

• we overestimate our ability to control an external event even if it's random;
• we consider ourselves better, stronger, more solid than average; 
• we consider the occurrence of positive events to be more likely for us than for others (and vice versa for negative events).

When we learn new information, we integrate it more easily if it is in our favour rather than against us. For example, if an individual initially feels that he or she is at risk for a severe form of COVID-19 at 40% and is told that the risk is "only" 30%, that individual will find it much easier to remember the new figure than if he had been told that the risk was 50%.

This asymmetry in the treatment of information in our favour against information against increases our risk-taking in the face of COVID-19. This bias may lead us to believe, wrongly, that we are more protected than others against COVID-19 or less likely to be seriously ill due to COVID-19. The idea that SARS-CoV-2 is primarily a danger to others can lead to non-compliance with health instructions.

This bias may explain in part why countries in Europe and North America, among others, did not immediately feel concerned about the epidemic when it was spreading to China at a time when there was considerable human flux worldwide.

Finally, this bias surely is partly responsible for a certain laxity in the respect of barrier gestures during the summer. Despite a strong spread of the virus, severe forms were less frequent, which could wrongly suggest that the coronavirus was less virulent. The decrease in severe forms this summer was actually related to the (younger) age of the infected persons and the fact that summer life outside reduced the infectious viral load and not to the decrease in the virulence of the coronavirus. See the question The severity of the COVID-19 disease. Does it depend on the dose of virus received?

4. Confirmation bias

This bias leads to an increased sensitivity to the elements that confirm our beliefs or hypotheses, and a reduced sensitivity to those that invalidate them. 

For example, if a person thinks that wearing a mask is not helpful in the fight against the spread of the SARS-CoV-2 coronavirus, he or she will tend to be preferentially sensitive to facts that point in that direction and not to consider facts that prove otherwise. Consequently, s/he may focus on the fact that there have been as many deaths in one country that has made the wearing of masks mandatory as in another that has not, without taking into account other potential explanatory factors such as population density, cultural habits, age of the population, or co-morbidity factors. 

5. Availability bias

Our estimate of the probability of an event is biased by the ease with which its occurrence comes to mind. If an event comes easily to mind, we estimate it as probable, unlike an event for which we have no examples in mind. In the case of COVID-19, if we do not personally know of any sick people, we risk underestimating the prevalence of the disease and respecting health instructions much less.

The underestimation of the risk for people who have not experienced severe illness is particularly relevant today during the COVID-19 pandemic. Low-risk individuals, referred to as "young and invincible", frequently underestimate their possible contribution to the risk of disease transmission and its impact on the population because there may be no severe cases in their environment.