[Press & Publications] Piecing together the evidence [Psychology]



March 2000

No 56

Psychologists investigate new techniques of identifying criminals

Researchers from the Face Processing Research Group (University of Leicester, the Open University and University of Westminster) have pioneered a new system of compiling E-FITs of criminal suspects.

The technique could help police to better identify criminals - based on the improved E-FITs arising from the technique.

The new 'jigsaw' method seeks to modify the existing system used by most police forces by reducing to a minimum the memory interference that distorts a witness's vision of a suspect.

Instead of compiling an E-FIT on the basis of a complete all-embracing description that is then fed into a computer, 'jigsaw' will build up an image piece by piece using only descriptions that the witness has supplied.

Dr Richard Kemp, senior lecturer in forensic psychology at the University of Leicester, said: "What we are actually seeking to do is to improve the performance of creating an E-FIT.

"Under the current E-FIT system, as a witness gives a description of, say, the eyes, the computer chooses the best set of eyes for you, and so on. The witness does not see the face until it is completed.

"What we are saying is that you can improve things by allowing witnesses to watch the face build up. However, as it is important that the features are always seen as part of a face, a cartoon outline and features are used to replace those elements that have yet to be described."

Dr Kemp said the new technique, which is being developed in conjunction with Jim Turner of the University of Westminster, had advantages over the current system because of growing knowledge about the way memory works: "Our memories are not photographic processes. People distort things, even forget things.

"Psychologists have been working with the police for quite a while to try and improve those memory processes. Changing the way police conduct interviews to elicit information for the E-FITs aims to get the best information possible out of the witness."

In addition, Dr Kemp and his team are investigating the benefits of using the evidence of multiple witnesses to create a more accurate image of a suspect's face.

"Police guidelines presently permit the evidence of only one witness in the compilation of an E-FIT. But the police have been interested for some time in what the effect would be of having multiple witnesses.

"It could be that seeing four E-FITs would confuse people. What we found was that it was actually better to show more - seeing the variation helps people."

  • Additional research conducted by the Face Processing Research Group is investigating whether running identity parades on video is better than doing them live. The team has found that witnesses can be so overawed by a live parade that they spend only a few seconds on each face. Using video, they can pause and rewind as much as they like. In tests, this led to 100% of volunteer witnesses making a correct identification. When the target was not present, again 100% of the volunteers said they did not recognise him in the parade.
  • NOTE TO NEWSDESK: For more information, contact Dr Kemp on 0116 252 3658.

    Please use correct title - UNIVERSITY OF LEICESTER - in any report.


    [Leicester University] [*] Administration [*] Press & Publications
    Information supplied by: Barbara Whiteman
    Last updated: 27 March 2000 16:01
    University Administration Web Maintainer

    This document has been approved by the head of department or section.
    If you are an authorised user you may edit this document through your Web browser.