US mini logoHome | A-Z Index | People | Reference | Contact us

Press release

  • 18 October 2007

Why the innocent get picked in ID parades

Two simple changes to the way police run identity parades would improve their accuracy, says University of Sussex psychologist Dr Dan Wright.

Dr Wright, whose recommendations have been submitted by the British Psychological Society (BPS) in response to a Home Office consultation exercise, points out that the current procedure leads to 20 per cent of witnesses choosing innocent people, which can lead to false imprisonment.

Based on research in the US and the UK, Dr Wright suggests using "double blind" video identifications where possible, so that the person directing the "parade" is not aware of the identity of the suspect. The procedure, already commonly used in medical research trials, would eliminate the possibility of the administrator inadvertently giving biased, non-verbal cues.

Additionally, measuring how confident a witness feels over their choice immediately after the parade, but before the identification of the suspect is revealed, would give a more reliable outcome.

Dr Wright, a member of the Memory and Law Working Party of the BPS, explains: "Police, judges and jurors use the confidence of witnesses to weight the reliability of their identifications. The problem is that a witness's confidence is affected after they are told either they identified the suspect or that they failed to identify the suspect. It is not possible to tell if someone is confident because they are accurate or just because they identified the suspect."

Dr Wright's recommendations were made as part of the BPS Submission in response to the Police and Criminal Evidence Act (PACE) review consultation exercise. "This consultation exercise provided the opportunity to make at least two procedural changes that are widely agreed to be beneficial by researchers in the area," said Dr Wright. "Neither of these changes should increase police work, but both will improve the reliability of identifications."

Notes for editors

  • The submission is available via,  and the summary of responses via
  • University of Sussex Press Office, Jacqui Bealing and Maggie Clune, tel:  01273 563326, email:


Useful links

Site maintained by: Web team Disclaimer | Feedback