12 January 2015 17:09

A new way to identify liars with over 70% accuracy has been developed by an international research team.

Researchers at Lancaster, Cambridge and Utrecht have made a breakthrough in developing a method that monitors a suspect’s full-body movements for signs of guilt.

The research, led by former Lancaster PhD student and Cambridge researcher Dr Sophie van der Zee, is based on the premise that people fidget more when they lie.  Using an all-body motion suit – the kind used in films to create computer-generated characters – the research team were able to pick this up this and identify liars with over 70% accuracy.

Professor Paul Taylor of Security Lancaster, who worked on the research, said: “This method is effective because of its sensitivity to subtle changes in behaviour that are difficult for human judges to detect.  Combine this robust measurement with effective questioning techniques, and the accuracy of this approach may improve still further.”

The experiment carried out by Van der Zee, Taylor, and their colleagues Ross Anderson of Cambridge University and Ronald Poppe of Utrecht University, involved 180 participants at Lancaster University lying and telling the truth about two experiences. 

The research suggests motion capture may be at least 10 per cent more accurate than the polygraph, which is widely used in the US for criminal cases.  The team’s finding is one from a line of research being conducted into alternative technologies and interpersonal markers of deceit.

The findings are to be published at an international conference on system sciences in Hawaii.