How Our Existing Beliefs About Vaccines Lead Us To Misinterpret New Information

How Our Existing Beliefs About Vaccines Lead Us To Misinterpret New Information

As Covid vaccines have continued to be rolled out across the world, ‘vaccine hesitancy’ has become the subject of growing media interest.

In some countries, surveys show a remarkable reduction in vaccine hesitancy. A recent survey of the UK population shows that 93% of people would accept a vaccine in the UK, up from 78% before Christmas.

But in other parts of the world, and in particular communities even within the UK, vaccine hesitancy continues to be a real issue. Recent polls suggest that nearly four in ten people in France, and close to a quarter of Germans would not get vaccinated if offered.

In a series of studies run in partnership with the global strategic communications agency Finsbury Glover Hering, we set out to find out why different people might have such different views about the safety and effectiveness of the Covid vaccine. And it turns out that a big part of the answer lies in people’s prior beliefs and expectations.

We asked 4,000 people in the US and the UK to look at some information regarding a new Covid vaccine. These people were a representative sample of both populations, and therefore contained people who were pro-vaccines, anti-vaccines, and everything in between.

All participants were shown a summary table (see below) pertaining to a recent (fictional) vaccine trial. Some of them were randomly allocated information and data showing the vaccine to be effective. Others were shown the same data table, but with the outcomes flipped: indicating that the vaccine was ineffective. We then asked participants to answer one simple question: based on the data in the table, was the vaccine effective or not?

What we found was that people who were vaccine hesitant systematically failed to get the questions right when they showed the vaccine was effective. But systematically answered the questions correctly when the data showed the vaccine was ineffective.

In behavioural science terms, this is classic ‘confirmation bias’. These people were seeking out and re-interpreting information so that it conformed with their existing beliefs.

But the really interesting finding, for those of us who might assume that pro-vaccine individuals are more enlightened and objective, was that pro-vaccine participants suffered from exactly the same bias. But in reverse. They systematically answered correctly when this data supported their existing views. But tended to get the question wrong when the data showed the vaccine to be ineffective.

What this means is that the provision of objective information cannot be relied upon to change beliefs or behaviour. And in some instances can make matters worse by reinforcing existing beliefs.

This is why governments around the world are – rightly – thinking about how to reach people and groups who might be vaccine hesitant through new channels, including campaigns fronted by individuals that different communities can relate to. And it is why we need to be careful when conveying information about vaccine rollouts being paused, when we think we are just stating matter-of-fact scientific information.

More generally, as we argue in the paper on risk communications with Finsbury Glover Hering, we need to understand that when communicating risk, information alone rarely changes or strongly influences people’s beliefs. If we want to achieve behavioural change, we need to consider who is conveying the information; the way it is framed; and its emotional and social context.

If you want to find out more about this and other studies we have conducted to test the communication of risk, our paper is available here.