The Open UniversitySkip to content
 

Facial Recognition from Identification Parades

Havard, Catriona and Memon, Amina (2012). Facial Recognition from Identification Parades. In: Wilkinson, C. M and Ryan., C. eds. Cranofacial Identification. Cambridge: Cambridge University Press, pp. 86–100.

URL: http://www.cambridge.org/gb/academic/subjects/life...
Google Scholar: Look up in Google Scholar

Abstract

Identification parades are one of the most common means of identifying a perpetrator of a crime and can be powerful evidence in securing convictions in criminal cases. In an identification parade (also known as a line-up) a suspect is placed amongst a number of similiar-looking people (foils) and the task for the witness is to either select the person they recognise as being the culprit, or state the culprit is not there. Unfortunately, witnesses do not always correctly identify the culprit from a line-up and in some cases innocent people are wrongly identified. This issue has been investigated by the Innocence project, who at the time of writing, have been involved in 251 exonerations based on DNA evidence in the USA. Of these cases approximately 75% of those convicted were cases of mistaken identity (Innocence Project at: http://www.innocenceproject.org/). Several similar organisations have developed worldwide such as the UK Innocent Network, the Australian Innocence Network and the Innocence Project New Zealand, all with the aim to overturn convictions of those who have been wrongfully imprisoned. One real life case of someone who was exonerated by the Innocent Project is Calvin Willis; he was wrongly accused of rape and served 22 years before DNA evidence proved he was innocent. Even though one of the police reports stated the victim did not see her attacker's face, she was shown a line-up and said she was told to pick the man without the full beard. She later testified that she did not choose anyone, although the police said she picked Willis. Willis's name only came into the investigation because the victim's neighbours had mentioned Willis's name when discussing who might have committed the crime. Even with all these flaws in the investigation the jury still convicted Willis and sentenced him to life (for more details of this case see Innocence Projects webpage at: http://www.innocenceproject.org/Content/297.php). This illustrates how powerful eyewitness evidence can be in the courtroom, even when there are obvious flaws in the way it has been obtained.

This chapter will explore a number of issues that can influence how accurately witnesses make decisions when viewing a line-up. These issues can be split into two separate strands: system variables and estimator variables (Wells, 1978). System variables are factors that are under control of the police conducting the parade, and include the method of foil selection and the identification format e.g. photo or video line-up. They also include whether the line-up presentation is sequential or simultaneous, instructions given to the witness, and whether they view a mugbook or make a composite prior to viewing a line-up. Estimator variables are factors associated with the witnesses and their view of the crime and are not under any control from the police or the judicial system. These include the witness's age and eyesight, how long they saw the culprit for and also whether the culprit was the same age or ethnic background as the witness. We will examine how system and estimator variables can influence facial identification from line-ups, but first the method used to study the factors influencing eyewitness identification accuracy will be described.

Item Type: Book Section
Copyright Holders: 2012 Cambridge University Press
Academic Unit/School: Faculty of Arts and Social Sciences (FASS) > Psychology
Faculty of Arts and Social Sciences (FASS)
Research Group: Harm and Evidence Research Collaborative (HERC)
Item ID: 47861
Depositing User: Catriona Havard
Date Deposited: 21 Nov 2016 09:26
Last Modified: 07 Dec 2018 10:46
URI: http://oro.open.ac.uk/id/eprint/47861
Share this page:

Actions (login may be required)

Policies | Disclaimer

© The Open University   contact the OU