Book Read Free

Handbook of Psychology of Investigative Interviewing: Current Developments and Future Directions

Page 44

by Ray Bull, Tim Valentine, Dr Tom Williamson


  -

  Klavehn, 2000 for a review). Applying this theoretical framework to eyewitness

  identifi cation suggests that it may be benefi cial in future research to explore

  use of instructions to witnesses that encourage a recollection judgement and

  discourages identifi cation based on a feeling of familiarity.

  Factors that a ffect e yewitness i dentifi cation

  Crime scene investigators take great care to avoid contaminating a crime scene

  by introducing new trace evidence. The memory of an eyewitness should be

  treated similarly, as part of the crime scene. Investigating offi cers should take

  care to ensure that they do not distort a witness ’ s memory, but instead use a

  sensitive and fair procedure to obtain identifi cation evidence of the highest

  quality. An investigating offi cer has control over many aspects of an eyewit-

  ness ’ s interaction with the investigation. For example, an investigator may

  develop a strategy to interview a witness, decide which witnesses will be asked

  to make an identifi cation, selecting an appropriate method, brief the witness

  and control the information provided to the witness after an identifi cation

  procedure.

  The selection and design of identifi cation procedures can have a strong

  impact on the reliability of the eyewitness evidence obtained. A poorly designed

  procedure might distort a witness ’ s memory, potentially leading to a mistaken

  identifi cation of an innocent suspect or a missed opportunity to collect

  Recent Developments in Eyewitness Identifi cation Procedures

  227

  identifi cation evidence against a guilty suspect. Psychological science can

  eva luate identifi cation procedures. For example, the selection of foils for a

  line - up, the instructions given to witnesses and prior viewing of photographs

  have all been shown to have a marked effect on the reliability of eyewitness

  identifi cation.

  Selection of f oils

  The PACE code of practice specifi es that the foils for line - ups must be selected

  to ‘ resemble the suspect ’ . This is known as a suspect - resemblance strategy.

  Luus & Wells (1991) argued that a better strategy is to select foils who match

  the witness ’ s description of the culprit. It is reasonable to assume that the

  witness can remember the description that he or she gave to the police and

  may expect to identify somebody who matches their description. Therefore,

  the witness may be inclined to disregard any foils that do not match their

  description, or conversely pay special attention to anybody who is a better

  match to their description than the rest. Luus & Wells suggested that it does

  not introduce a bias against an innocent suspect if line - up members differ on

  a feature that was not mentioned in the original description. Differences in

  the facial features of line - up members will help a witness who has a reliable

  memory to distinguish the culprit from the foils. If the suspect is not the

  culprit, he or she is no more likely than anybody else to be mistakenly identi-

  fi ed by a feature not mentioned in the description because the witness has not

  seen the suspect before. All persons in a fair line - up should match the witness ’ s

  description of the culprit.

  Sometimes a witness may not mention the sex or race of a person, or may

  neglect to say that somebody did not have a beard or was not wearing glasses.

  This may occur because the witness assumes a default value (Lindsay, Martin

  & Webber, 1994 ). When constructing a culprit - description line - up, account

  must be taken of default values. Alternatively, if a witness ’ s description is very

  vague or if the suspect does not match the description, it may be necessary to

  resort to a suspect - resemblance strategy. The important principle is that the

  suspect should not stand out in the line - up.

  Different foils may be required for a culprit

  - description line

  - up for

  each witness, because their descriptions of the culprit may differ. Therefore,

  use of culprit - description line - ups would require additional police resources.

  In the UK current practice requires only a single line - up to be used for all

  witnesses, although the position of the suspect may differ across different

  witnesses.

  Empirical support for the superiority of a culprit

  - description line

  - up is

  mixed. Wells, Rydell & Seelau (1993) and Juslin, Olsson & Winman (1996)

  reported more correct identifi cations from culprit - description line - ups than

  from suspect - resemblance line - ups when the culprit was present in the line - up,

  and no signifi cant difference in the number of mistaken identifi cations from

  culprit - absent line - ups. In contrast, Lindsay et al. (1994) , and Tunnicliffe &

  228

  Handbook of Psychology of Investigative Interviewing

  Clark (2000) did not fi nd any statistically signifi cant advantage for culprit -

  description line

  - ups over suspect

  - resemblance line

  - ups. Use of culprit

  -

  description line - ups in the context of video identifi cation is discussed below.

  Instructions to w itnesses

  A witness may believe that they have been invited to attempt an identifi cation

  because the police have good reason to believe that the suspect is guilty.

  Therefore, a witness may assume that it will help the police if they identify the

  suspect. It is important that the instructions to the witness should emphasize

  the possibility that the offender may not be in the line - up. The PACE code

  of practice includes the instruction that the person whom the witness saw ‘ may

  or may not be present ’ . Furthermore, the witness must be instructed that ‘ If

  you cannot make a positive identifi cation, you should say so ’ . Instructions that

  do not point out that the culprit may not be in the line - up are regarded as

  ‘ biased ’ (e.g., ‘ Can you identify the man who assaulted you? ’ ). A meta - analysis

  of 18 studies showed that when biased instructions are given, witnesses are

  more likely to make an identifi cation, whether it is correct or incorrect. Biased

  instructions increase the likelihood of an innocent suspect being identifi ed

  from culprit - absent line - ups (Steblay, 1997 ).

  Blind a dministration of l ine - u ps

  ‘ Blind ’ is used in the sense that the line - up administer does not know (i.e. is

  blind to) the identity of the suspect. The procedure is often referred to as

  ‘ double - blind ’ , meaning that both the witness and the line - up administrator

  are blind to the identity of the suspect. A double - blind procedure should be

  used to prevent any inadvertent infl uence on the witness. Such infl uence can

  be very subtle and entirely unconscious. For example, the administrator may

  look at the witness when the suspect ’ s image is being viewed, or be more likely

  to accept a tentative identifi cation only if it is of the suspect. Currently, the

  PACE code of practice does not require double - blind administration.

  There is little direct evidence on the double - blind administration of line -

  ups. However, the effects of experimenter - induced i
nfl uence are well estab-

  lished in a broader context in psychological science (Harris

  & Rosenthal,

  1985 ). For this reason a double - blind procedure is an essential feature of clini-

  cal drug trials. The essential point to appreciate is that double - blind adminis-

  tration of identifi cation procedures removes any possibility of that the witness

  may have been infl uenced. Therefore, double - blind administration protects the

  police from malicious accusation of bias and enhances the perceived integrity

  of the identifi cation evidence. Widespread use of photographs for identifi cation

  in the USA and of video in the UK greatly facilitates use of a double - blind

  identifi cation procedure.

  Recent Developments in Eyewitness Identifi cation Procedures

  229

  Number of f oils

  Video or live line - ups in England and Wales are required to have a minimum

  of eight foils. It is recommended by the US Department of Justice that photo

  line - ups have a minimum of fi ve foils and live line - ups have a minimum of four

  foils (Technical Working Group on Eyewitness Identifi cation, 1999 ). These

  guidelines raise the question of how many foils is optimal.

  The similarity of foils to the culprit is more important than the abso-

  lute number of line

  - up members. A foil who is dissimilar to the culprit

  will be discounted by the witness. The ability to identify a target face

  from target - present and target - absent photo line - ups decreases with increa-

  sing size of the line - up (Meissner et al. , 2005 ). Thus the target person is

  most likely to be identifi ed from a one - person line - up (a show - up or con-

  frontation) and least likely to be identifi ed from a 12

  - person line

  - up.

  This result is unsurprising because a line

  - up that offers more alternative

  responses provides a more diffi cult test of memory. It is only an identifi ca-

  tion of an innocent suspect that is a forensically relevant mistaken identifi cation

  (i.e., likely to lead to a miscarriage of justice). In a one - person target - absent

  line - up all mistaken identifi cations will be of the innocent suspect. Fewer

  mistaken identifi cations are likely to be of an innocent suspect in a large

  target - absent line - up than in a small line - up. In summary, there is no clear

  empirical guide to the optimal size of a line - up, but if foils are appropriately

  chosen, a larger line - up is likely to pose a more challenging assessment of

  eyewitness memory.

  Prior e xposure to p hotographs

  If the police have not identifi ed a suspect, the witness may be shown photo-

  graphs of people associated with similar offences (mugshots). This procedure

  differs from a line - up in that all the people are potential suspects. Therefore,

  any identifi cation will lead to that person being investigated. Later in the

  investigation the police may want to collect formal identifi cation evidence from

  a line - up, or may be required to do so by legislation. Would a subsequent

  line - up be biased against the suspect if the witness has previously seen their

  photograph in a mugshot album?

  Deffenbacher, Bornstein

  & Penrod

  (2006) provide a systematic review

  of the effects of mugshot exposure. They found that prior viewing of a pho-

  tograph of somebody who subsequently appeared in a line

  - up increased

  the probability of a mistaken identifi cation from the line - up. This effect is

  due to transference of familiarity from the photograph which is mistakenly

  attributed to having being seen at the crime scene. The effect is stronger when

  few mugshots were viewed (8 – 15 or fewer) than when more mugshots have

  been viewed. The effect is particularly strong if the person was mistakenly

  identifi ed as the perpetrator from the mugshot photographs. This is known as

  230

  Handbook of Psychology of Investigative Interviewing

  an effect of commitment to the earlier identifi cation. There was no ill - effect

  of showing photographs if none of the people seen appeared in the subsequent

  line - up.

  Video i dentifi cation and i ts a dvantages

  Since 2003 video technology has been used to replace virtually all live identity

  parades in England and Wales. A video line - up consists of 15 - second clips

  showing the head and shoulders of each line

  - up member. First, they are

  looking at the camera and then rotate their head to show both profi les, before

  looking back at the camera. The images are captured under standardized con-

  ditions. Each line - up member is shown sequentially, with a digit in the top

  left corner of the screen to identify each individual.

  Video offers a number of benefi ts compared to live line - ups:

  1. Video can dramatically reduce the delay before an identifi cation can be

  organized. A video line - up can be produced and transmitted via a secure

  network within two hours of request.

  2. The administration of the video identifi cation procedure can be arranged

  at the convenience of the witness rather than at the convenience of

  the suspect, because the suspect does not attend the video identifi ca tion

  procedure. This results in fewer identifi

  cation procedures being

  cancelled.

  3. A large database of video clips (in excess of 20,000) is available, providing

  more foils for selection. This makes it easier to select foils that resemble

  the suspect in appearance.

  4. Video is less threatening to victims, who no longer have to attend an

  identifi cation suite where, for example, their attacker may be physically

  present.

  5. Video equipment can be taken to a witness who is unable to attend the

  police station. For example, a victim of a violent attack may be able to

  view a video line - up from a hospital bed or a witness could view a line - up

  anywhere in the world.

  6. The suspect no longer has the opportunity to change their appearance if

  their image was captured for the video identifi cation when they were fi rst

  detained.

  7. A video line

  - up does not require the continuing co

  - operation of the

  suspect. Once the suspect ’ s image has been captured for a video line -

  up, the suspect

  ’ s co

  - operation with the procedure is no longer

  required.

  Video identifi cation has had a considerable impact on the investigative

  process due to development of legislation and technology. In certain instances

  when there is a suspect known to the police but who is unavailable, identifi ca-

  Recent Developments in Eyewitness Identifi cation Procedures

  231

  tion offi cers may hold an identifi cation procedure in advance of any arrest using

  any suitable still or moving image of the suspect. This same principle of using

  historically held imagery may be applied to a person suspected of committing

  a crime many years ago and whose appearance has changed ( R v Folan , 2003 ).

  Forensic advances in the use of DNA evidence and an ever - increasing database

  of suspect images means that this aspect o
f crime investigation is set to increase.

  Overall, the speed of process that video identifi cation provides has, in many

  instances, hastened the outcome of an investigation where identifi cation is

  deemed an issue. It is not unusual, particularly where serious crime is involved,

  for the identifi cation procedure to be conducted whilst the suspect is in

  custody. A further feature of video technology is the ability to cover or replicate

  distinguishing features on the faces of the suspect and/or the foils (e.g. scars,

  tattoos etc.) to ensure a procedure complies with legislation. This issue is

  discussed in more detail below.

  The i ncreased u se of e yewitness i dentifi cation e vidence

  Devlin (1976) ascertained that 2,116 identity parades were held in England

  and Wales during 1973. PACE resulted in an increase in the use of identity

  parades, because it gave suspects the right to an identity parade if identifi cation

  is disputed. Slater

  (1994) reported data from 24 of 52 UK police forces

  showing a sharp rise in the number of identity parades organized during the

  period 1990 – 93. In 1993 the 24 police forces who reported data organized

  13,652 identity parades. The introduction of video together with further leg-

  islative changes, which has made it easier to organize identifi cation procedures,

  has resulted in a further sharp increase in demand. The best estimate for 2006

  suggests that the number of video identifi cation procedures had reached a

  minimum of 80,000 annually. Paradoxically, the courts now rely more on

  eyewitness identifi cation than they did in the 1970s when Devlin analysed

  eyewitness identifi cation evidence.

  Research on v ideo i dentifi cation

  The format of presenting a line - up (photographs, video, live) and manipulation

  of the richness of cues available (e.g. stills, moving images, people walking)

  has a surprisingly small effect on identifi cation accuracy. A possible reason is

  that the face is the most reliable way to recognize somebody and the face can

  be suffi ciently well perceived from a good quality still photograph. Therefore,

  relying on cues such as gait, build or colour images would add little extra

  benefi t. Reviewing the literature, Cutler, Berman, Penrod & Fisher concluded:

  ‘ With respect to current practices, the conservative conclusion is that, based

  on available research, there is no reason to believe that live line - ups, videotaped

 

‹ Prev