Investigating the biases in the perception of human behaviour using immersive virtual reality

Project: Research project (funded)Research

Project Details


Accurate perception of the actions and behaviour of other individuals is essential for making correct decisions on how to respond and interact and is therefore critical to functioning successfully in our complex social world. Incorrect interpretation of the behaviour of other individuals can have very profound consequences, causing difficulties in social interactions resulting in anxiety and stress. In addition, individuals with schizophrenia and autistic spectrum disorders often make incorrect judgements of human actions due to their disorder. For individuals that make critical behavioural judgments as part of their job, (e.g. police monitoring of CCTV, security service surveillance of terrorism suspects) incorrect interpretation of human behaviour can have very severe consequences. Establishing the causes underlying incorrect behavioural judgements, and working towards the possible mitigation of these problems, is therefore of immediate importance.
Our perception of the world can be biased over the short term by visual processes and is not always accurate. There is well documented evidence that prolonged viewing (>5seconds) of relatively simple stimuli (colours, oriented lines, moving dots/bars) results in "adaptation" causing "after-effects" where perception is profoundly biased. Recently we have demonstrated that adaptation can occur after viewing human actions and behaviour. The focus of this project is therefore to characterise the inaccuracy and biases in our perception of human actions and behaviour due to visual adaptation, which may elucidate the function of underlying mechanisms operating in natural social environments.
Studies to date have used standard psychological testing under laboratory conditions. It is not entirely obvious how these results translate to human performance in more naturalistic conditions. This project will therefore combine the rigour of precisely controlled and established psychological testing with a very high resolution 3-dimensional full sized presentation of actors in naturalistic environments in a state-of-the-art "Virtual Reality" environment. This novel combination of techniques, only now realizable, allows us to investigate human perception and brain mechanisms in realistic environments not available using traditional psychophysical or neuroimaging techniques. We will compare results obtained under standard laboratory conditions with those from the virtual reality environment to provide important new evidence about how perception and brain function assessed using traditional psychological techniques is applicable to simulated "real life" perception.
We will profile the strength and duration of the effect of visual adaptation in causing biases in the recognition of actions, perception of human behaviour and judgments we make about other people's expectations. These results are important because they quantify the extent to which action and social perception is biased by immediate perceptual history, and in addition provide information on the degree to which adaptation influences the different stages of visual processing.
Experiments will be carried out to determine how our perception of the trustworthiness of other people can be biased by immediate prior experience. We will also measure the extent to which the perception of trustworthiness of individuals viewed on CCTV screens is biased as well as how Police officer's judgments of trustworthiness are biased by their immediate prior visual experience. Together these results will tell us about the mechanisms underlying the perception of trust from non-verbal behaviour and provide important information on how to mitigate biases in the perception of trust in security professionals.
This project brings together rigorous visual experimentation and state of the art visualisation technology to examine and improve our fundamental ability to make sense of the behaviour of other people in natural environments.

Layman's description

Accurate perception of the behaviour of other individuals is essential for making correct decisions on how to respond to and interact with them and is, therefore, critical to successful social functioning. Perception, however can be profoundly biased over the short term (<10s) by our immediate prior experience, leading to incorrect judgments of our external social world. This project aims to investigate how our judgments of human behaviour can be biased by our immediate prior experience, and how security professionals who have to make decisions about human behaviour under “mission critical” situations are influenced by short term changes or distortions in their perception of the current visual input. We will exploit these distortions in perception to investigate the mechanisms underlying social perception within simulated naturalistic situations by employing a novel use of immersive virtual reality techniques.
Effective start/end date2/07/1131/05/14