Sometimes we need to learn from others

When you’re trying to decide what to order from a familiar menu, you’ll most likely draw from your own experience.

But if you haven’t been to that café before, it might help to watch what others choose – and that’s what people are more likely to do, according to a study published in the journal Science Advances.

The research shows that each mode of learning – direct or social – involves different but linked neural pathways in the brain.

“Humans learn from their own trial-and-error experience and observing others,” write Lei Zhang and Jan Gläscher, from Germany’s University Medical Centre Hamburg-Eppendorf. “This concerns both big and small decisions alike.”

Studies have tried to pinpoint brain regions associated with social influence on decision making, but Zhang and Gläscher say how this interacts with direct learning has rarely been considered.

They recruited 185 volunteers to take part in a multi-stage gambling experiment in groups of five, of which 39 – one from each group – had their brains scanned using functional magnetic resonance imaging (MRI).

Each volunteer was presented with two abstract symbols and was told that one would lead to cash rewards in the long run.

In each round of the experiment, people first chose one symbol and then watched which symbol the other four had selected. Then they could decide whether to stick with their original choice or choose the other one, before being given a reward based on their decision.

“This way, we enable real-time interactions among the volunteers, which greatly enhances ecological validity,” says Zhang.

In a twist, the symbol that was linked to a reward kept changing. At first, one symbol earned a reward 70% of the time, but after a few rounds it was only rewarded 30% of the time – and these changes took place multiple times.

“This so-called reversal learning paradigm will create uncertainty for volunteers so that they will always need to learn and relearn to gain more outcomes,” explains Gläscher.

“In particular, when the reversal just happened, some people in the group may pick it up faster than the others, and if so, the others could combine this social information into their own decision-making processes.”

The volunteers switched more often when others made different choices, but their second choice – after watching others – better reflected the reward structure than the first.

To unravel this, the researchers modelled people’s behaviour and found they used different strategies for direct learning and social learning.

“At the beginning of each round, the volunteers were combining their own direct learning experience and social learning experience to guide their choice,” says Zhang, “whereby direct learning follows a simple reinforcement learning algorithm, and social learning is instantiated by tracking the others’ reward history.”

Brain scans reflected these differences, showing that direct learning recruited a region called the ventromedial prefrontal cortex and social learning a region called the anterior cingulate cortex, which interacted in an area called the striatum.

“These indicate an integrated brain network supporting social influence in human decision making,” says Gläscher.