New study challenges current thinking on how attention changes with social context

The study tested three conditions where people observed footage of two women in a waiting room, with different groups told that they were watching a live webcam and either told they would or would not meet the women afterwards whilst others were told that the video had been pre-recorded.

Explaining what they expected to see, Dr Gregory said: "We thought that when participants believed that they would be meeting the people in the scene, they would have their attention drawn towards the faces of those people more readily, and look where they looked more often, than the other two groups as the people would be most socially relevant to the participants."

"We also expected that in the condition which was least like real-life, when people thought the scene was pre-recorded, they would look least at the faces of the actors and follow their gaze direction the least.

"However, we found the complete opposite. Regardless of whether they thought they would meet the people in the scene, when participants thought they were watching a live webcam they seemed to avoid looking at the faces of the people and hardly followed their direction of gaze at all even though the people in the scene could not see the participants. When participants thought the scene was pre-recorded, they looked at the faces and followed gaze direction of the actors much more.

"Perhaps what we think we know about the way we view other people is wrong. As soon as viewing behaviour is measured within a genuinely social context, the way we look at people changes, and rather than having our attention drawn towards them, we actually seem to avoid looking at those people's faces.

"This likely reflects the complex interplay of factors which are present in a real social scenario, which are absent in most experimental studies, such as adhering to social rules and norms, or thinking about lots of different things at once, which causes us to look less at people than when we view pictures of them in the lab.

"Psychologists need to start taking this into account in their research, which is really very rare at the moment to make sure that what we find out from our experiments can actually be applied to real-life behaviour. If it can't then the value of it must be reassessed."

Nicola Gregory's research provides fresh insight to this research field, taking place in a more natural and social context unlike the older studies. It also refutes previous work suggesting that people predominantly look at faces and automatically shift their attention in the direction that other people are looking.

The study is published online in science and medicine journal PLOS ONE, a non-profit organization of scientists and physicians committed to making scientifically-useful literature freely available.