Computer simulation uncovers possible reasons for prejudices

19 April 2009

A study by Professor Richard Eiser, from the Department of Psychology at the University of Sheffield, has shown that computer simulation can indicate why people´s prejudices are resistant to change.

The study used computer simulation to look into why negative beliefs and prejudices resist correction. By giving a computer programme a problem to solve, the study determined how it responded to each attempt, and concluded the brain can process information in a similar way.

In the simulation, good objects were given a negative bias, and bad objects were given a positive bias. A simulation was then run to see whether the system would overcome its biases to realise that the good objects were actually good and bad objects bad.

The study discovered that biases guided whether objects were approached or avoided. The bad objects given a positive bias were approached regularly, so the system learned that these objects were actually bad (identifying bad objects with a 100 per cent success rate).

The good objects with a negative bias were avoided; so the system never learned that it was wrong (with the good objects that were given a negative bias only being identified at a 32 percent success rate).

Professor Eiser said: "We can take these findings as an illustration of the thought processes around prejudiced thinking such as racism or homophobia. Prejudices mean people avoid contact with disliked groups and so their negative beliefs are never proven wrong."

He continued: "This computer simulation successfully demonstrated learning processes, which may be part of the story of why our prejudices are resistant to change."

The study has been published today (20th April 2009) in the British Journal of Psychology.

Logo Careerjet