What makes people believe their emotions instead of science or statistics?
By analyzing a large number of studies, a group of scientists believe they’ve identified the reasons why people may be persuaded by emotional or personal stories over statistics .
In Short: When an issue doesn’t directly affect you or doesn’t seem serious, you are more likely to be persuaded by statistics. When you are directly invested in an issue, including one that you think is important or that directly affects you, you are more likely to be persuaded by anecdotes over statistics.
Why are statistics more persuasive in some cases, but not others?
While scientific evidence (with statistics!) can be very persuasive, we’ve all been in a situation where someone disagreed with scientific evidence because their friend’s cousin had a different experience. This person’s story or anecdote shouldn’t be more persuasive than statistics, but it often is, and a group of scientists wanted to figure out why this might be the case.
There is a lot of research on how people can be persuaded, but different studies often come to different, contradictory conclusions. In a new study, researchers decided to analyse a large number of previously-published (and unpublished) studies to see if there was a consistent trend. This type of study is called a meta-analysis, where the results of earlier studies are combined and the researchers try to see if they can see any statistical patterns. Meta-analyses can help to make sense of large amounts of data when different studies contradict one another.
For this particular case, the scientists reviewed 61 earlier studies that looked at why anecdotal evidence may be more persuasive in some cases, while statistical evidence can be more persuasive in other situations. Their work shows that when people are more emotionally-engaged in a topic, they are more likely to persuaded by anecdotes; when people are less emotionally-engaged, statistics are more persuasive.
The researchers used three different considerations to understand how emotionally-engaged a person may be with a topic.
threat severity: will there be a bad result if I make the wrong decision?
personal relevance: will the decision I make affect only other people, or will it affect myself?
health-related: these types of decisions often involve more uncertainty and fear
Based on this, they found that if a person is emotionally-engaged, based on one or more of these criteria, they are more likely to be persuaded by an anecdote, and less likely to be persuaded by statistics.
While this research is promising, it is important to know that meta-analyses are not perfect (nor is any study). They typically look at things that are in common between studies. In this case, if the researchers could not look at threat severity, personal relevance, or health-related considerations in previous studies, they did not include that study in their meta-analysis. Meta-analyses are therefore better at seeing trends in studies, rather than looking at every single possibility.
As scientists around the world race to develop a coronavirus vaccine for COVID-19, there will still be people that oppose life-saving vaccines, even during a pandemic. The meta-analysis by Freling and their colleagues is important for understanding why people may be less likely to believe science than a story they heard from a friend, and be useful for helping public health officials communicate the effectiveness and importance of vaccines.
The fully study can be found here, but it is behind a pay-wall. I’ll write a blog post in the future on how to access published studies without needing to pay for access.