Figure from Bethlehem et al showing the staged crash & observers locations
Social and behavioural, Ethics and morality

Staged bike crash tests whether empathic people are more altruistic

Study stands out because it was conducted outside of the psychology lab.

02 November 2016

By Christian Jarrett

You’re walking to work and spot a cyclist on the ground, next to his upturned bike, wincing in pain. Do you go and help? Of the many factors influencing your decision, psychological theory suggests that among the most important is your levels of empathy. If you feel the cyclist’s pain and misfortune, you’re more likely to be motivated to help. This might sound obvious, but there has been surprisingly little research to test whether measuring someone’s empathy levels in a questionnaire actually predicts the likelihood that they will show real-life altruism. That’s what Richard Bethlehem and his colleagues have done for a new open access study in Social Neuroscience, in which they staged a bicycle accident along a university footpath. The results provide some of the first evidence that empathy is correlated with altruism “in the wild”. 

Like secret agents on a surveillance mission, the researchers placed observers in two discreet positions opposite and after the staged bike accident scene (in which the cyclist was sitting on the ground, wincing and rubbing his ankle). The first observer took notes on all passersby approaching the crash, and signalled to the second observer, positioned in a concealed location after the crash, whether the next person to pass the scene was physically unimpaired and on their own, making them eligible for the study.

The second observer then noted if an eligible passerby helped the cyclist or not (if approached, the cyclist said he was fine and just resting) and, either way, she approached and asked this person to take part in a memory study – this was to conceal the true aims of the research. If they agreed to take part, they became a study participant, and the observer then asked this person questions about memorable features of his or her journey, and then took his or her email address for sending of questionnaires to be completed later. These questionnaires tapped empathy levels and autistic-like traits.

Of the 1067 eligible people who walked by the crash scene, 55 subsequently agreed to talk to the second observer and take part in the study. Of these participants, 29 per cent had stopped to help the cyclist (compared with just 7 per cent of the entire sample of 1067). Analysis of the participants’ later questionnaire scores showed that empathy scores were correlated with real-life altruism – that is, the good Samaritans scored much higher on empathy than the non-helpers (average score 56/80, versus 20/80).

Some earlier research has suggested that people with more autistic traits are less likely to show altruistic behaviour, but in fact the autism questionnaires scores were not correlated with helping behaviour, although bear in mind that scores were skewed to the low end. But note also, one participant did score highly on the autism questionnaire (and it later turned out that he has a diagnosis of Autism Spectrum Disorder) and he was one of the passersby to help the cyclist.

This study stands out because it was conducted outside of the psych lab. “These types of real-life settings have become extremely scarce,” the researchers said. The findings suggest that most people do not stop to help a stranger, and that among the factors affecting our willingness to help – including the culture we are raised in, and how rushed we feel – our empathy levels remain an important influence.

“The implication of the present study is that within any institution (even perhaps extreme inhumane institutions such as those under the Nazi regime), there will be individual differences in how people within the institution respond, and that some of this variation in helping behaviour is accounted for by where on the empathy dimension the individual is situated,” the researchers said.

Readers might like to contrast these results with another paper published earlier this year by Paul Bloom (author of the forthcoming book Against Empathy) and others, which showed in a series of lab studies that people’s empathy did not predict their charitable giving, but their levels of concern did, prompting the researchers to argue that concern and empathy are two different things, and that concern, and not empathy, is the emotion that leads to altruistic behaviour.

Further reading

Does empathy predict altruism in the wild?

– Image shows figure from Bethlehem et al showing the staged crash & observers located at positions A & C.