Live, video, and photo identification: Which eyewitness identification procedure best facilitates recognition?

After observing a crime, eyewitnesses can provide law enforcement personnel with important information about the criminal's identity. This is often achieved by pointing out the criminal from an identity parade. The basic procedure for an identity parade is to have the witness inspect a group of people and judge whether one of them is the criminal, but the procedures used to administer identity parades vary substantially across jurisdictions. In countries like Australia and South Africa, parade members appear in person - similar to how identification procedures tend to be depicted in popular films. Live parades are also used in the USA and Canada, but identification from photographs (or "mugshots") is far more common. In the UK, recent legislation has led to the widespread use of video technology to present identity parades. Such disparity in procedures begs the question: Which one is best?

When considering the pragmatics of the different techniques, photo and video identifications are the clear favourites. Finding suitable people to appear in a live parade is not always easy, whereas the availability of large databases of photo and video images make these alternative forms of identity parades substantially more manageable to construct. Video and photo parades are also less expensive than live parades because the parade members do not need to be paid each time they appear.

Although the practical benefits of photo and video parades are clear, eyewitness scientists have yet to reach a consensus on which procedure is most likely to facilitate a correct identification decision. This is a huge societal problem because if a witness is unable to identify the criminal, investigators may not have the evidence required to secure a conviction and hold the criminal responsible. There is also a very real possibility that an eyewitness will mistakenly identify an innocent person. When this happens, the innocent person is at an increased risk of wrongful conviction. Wrongful convictions happen quite frequently and in a review of convictions that were proven to be wrongful in light of DNA evidence, mistaken identifications were present in over 70% of the cases and was the leading contributing factor (www.innocenceproject.org).

Sometimes false identifications are beyond the justice system's control, but decades of scientific research have demonstrated that eyewitness identification policies and procedures can have a significant impact on the reliability of eyewitness identifications. This is why it is crucial to know how identifications are affected by presenting the identity parade members in person, on video, or with photographs. One thing to consider is that live procedures give witnesses the opportunity to see the parade members in their entirety, whereas photo and video procedures typically only provide a head-and-shoulders view. All else equal, the ability to see the parade members from top to bottom should improve eyewitness identification performance. In practice, however, the logistical constraints of live identifications might wipe out any of these theoretical benefits.

The proposed research will provide a rigorous comparison of live, video, and photo parades, and it will also introduce a novel procedure that shows videos of the entire person. The novel procedure has been designed to retain the practical advantages of video parades, but also provide witnesses with the additional memory cues that are available for live parades. The people who are investigating crimes and setting eyewitness identification policy are currently facing a paradox: the most practical identification procedure may not produce the best identification outcomes. When the proposed research has been completed, the data will inform investigators and policy-makers on which identity parade modality produces the best identification outcomes.

The proposed research will compare three existing eyewitness identification systems and also explore the potential utility of a new system. This research would benefit all members of society who believe that a justice system should be both effective and fair. Everyone would benefit from identification procedures that hold the guilty responsible for their actions and protect the innocent from wrongful prosecution. In addition to these society-wide benefits, the following organisations and members of society are particularly likely to benefit: policymakers, law enforcement investigators, identification officers, eyewitnesses, innocent suspects, and taxpayers.

Policymakers need to know how the system that is currently in place compares with alternative systems. There is notable international variation in whether identity parades are viewed in person, on video, or with photographs. If one procedure is found to be more or less effective than the others, policymakers will have the information they need to decide whether the current system best serves its purpose.

Law enforcement investigators, who use parades to help establish whether a suspect is guilty or innocent, would benefit from knowing which identity parade modality yields the most informative evidence. If investigators are not using the most effective identification procedures, witnesses will make more mistakes and these mistakes could lead the investigation astray.

Identification officers, who administer the parades, could also benefit. In countries such as Australia and South Africa, live identity parades are the primary identification technique. Live parades present a number of logistical difficulties that may be avoided through the use of video or photographic images, but the case for transitioning to these more practical methods is difficult to make without conclusive evidence of their effectiveness.

Eyewitnesses could benefit in several ways. First, the implementation of the most effective identification procedure would facilitate the ability of eyewitnesses to identify the person who committed the crime and decrease the likelihood that they will misidentify an innocent person. If an eyewitness is unable to identify a guilty suspect, there is a reduced likelihood that the perpetrator will be held responsible for his or her criminal actions. Alternatively, if an eyewitness misidentifies a parade volunteer who is known to be innocent (i.e. a filler), the credibility of the witness is weakened. Worse yet, if the witness misidentifies an innocent suspect, there is an increased risk that an innocent person will be wrongfully convicted. If the innocent person is fortunate enough to be subsequently exonerated, the witness may have to endure feelings of guilt associated with sending an innocent person to prison. An additional consideration for eyewitnesses is that some procedures are more user-friendly than others. Eyewitnesses can experience a great deal of distress at an identification procedure, particularly if they are required to reencounter the perpetrator at a live identity parade. Data supporting a transition to video or photo parades would eliminate the need for this to happen.

Innocent suspects would obviously benefit from their reduced risk of mistaken identification and wrongful conviction.

Taxpayers may also benefit, as wrongful conviction cases commonly result in windfall payments to the exoneree in an attempt to compensate for the miscarriage of justice. Taxpayers may also benefit if a cost-effective procedure is adopted. Live parades are expensive because volunteers need to be paid to appear with the suspect. If the data support the use of video or photo parades, taxpayers in countries that use live identity parades would benefit from a transition to these more cost-effective techniques.