Love thy neighbor, unless they are The Weakest Link

Psychologists at the University of Lincoln who examined the behaviour of contestants on the BBC show found players were statistically less inclined to vote off the person stood next to them and more likely to target competitors stood further away.

The findings will be presented July 13 2012 at the Society for the Advancement of Behavioural Economics (SABE) Conference 2012, held in Granada, Spain.

The study, conducted by Dr Paul Goddard and undergraduates from the School of Psychology at the University of Lincoln, showed evidence of 'proxemic bias' in the players' decision-making.

This observation, that the physical space between people determines how they react to each other in certain social situations, drew parallels with findings from a classic but controversial social psychology experiment of the 1960s.

In a series of laboratory tests, Yale psychologist Stanley Milgram found people were more reluctant to administer a punishment they believed to be an electric shock to another person if they were located in the same room. When the 'punishment' could be delivered remotely from another room, the inhibition was reduced.

Modern day social psychologists recognise that TV quiz shows can provide a useful source of data for observational field studies because they avoid some of the ethical issues associated with laboratory experiments and are also effectively double-blind. The format of The Weakest Link is an example of 'forced choice decision-making' -- a context useful for understanding how subject-specific bias can influence choices.

The Lincoln researchers analysed voting patterns in the first round of 72 episodes of The Weakest Link, comparing how votes would be expected to fall based on probability, against how votes were cast in reality. They found contestants showed a strong reluctance to vote for their direct neighbour -- a pattern they called 'the neighbour avoidance effect'. This bias was stronger when the overall consensus about the worst player's identity was weaker.

Dr Goddard, Senior Lecturer in Lincoln's School of Psychology, said: "In the show contestants must make a choice about who is the worst player based on two very different sources of information. The primary and most reliable source comes from the game itself. If one player gets all their questions wrong, it's a fairly straightforward decision to vote them off. "The quandary for contestants arises when there is no clear consensus about who is the worst player, such as in rounds where several players get just one question wrong. In these circumstances contestants have to rely on a secondary source of information -- their own judgement. This is where bias can really come to the fore. "We found strong evidence of a 'neighbour avoidance effect' which supported our prediction that spatial proximity would influence players' decision-making."

Further analysis also showed evidence of gender bias in voting patterns -- women were significantly over-represented as the Weakest Link. Both male and female contestants were more likely to vote off a woman than a man.