Police Information

The Accuracy Of Eyewitness Information

Wed, September 19, 2001

Source: Mark R. Kebbell & Graham F. Wagstaff

FACE VALUE? EVALUATING THE ACCURACY OF EYEWITNESS INFORMATION Policing & Reducing Crime Briefing Note

Summary 

Eyewitnesses play an important part in many crime investigations. Hence it is important that the police understand how to assess the reliability of the information that witnesses provide. This paper provides a practical summary of current research for operational officers. Various factors have been identified that influence the quality of eyewitness statements, in particular: 

• The nature of the offence and the situation in which it is observed;

 

• The characteristics of the witness; and

 

• The way in which witnesses are interviewed. 

Whilst the police have no control over the first two of these, they can exert considerable influence on the way in which information is retrieved through, for example, interviewing techniques. 

Factors relating to the crime 

Many of the more critical factors, which influence the ability of eyewitnesses to give accurate information, were identified in the well known Turnbull Ruling (R. v Turnbull and others, 1977). These broadly take account of the conditions of the observation, the relationship of the witness to the event and certain aspects of the witnesses’ recollection of the event.  

Whilst the Turnbull Ruling provides good general guidance, it is important to be aware of exceptions. For example, victims of serious crimes can sometimes maintain accurate memories of an offence over a long period of time, as they tend repeatedly to go over the event in their minds, thus aiding storage in memory. Also, the presence of a weapon or threat of violence,

will increase the intensity of a witness’s experience, and raise stress levels. Stress will lead to a narrowing of witness focus and this is likely to concentrate attention towards the incident itself and away from ‘peripheral information’. 

Research has found that, generally, witnesses are more accurate with their descriptions of certain offender characteristics (such as sex, hair colour and distinguishing features) than others (such as age, height and specific race). The ability of witnesses to estimate correctly these characteristics will, however, be partly determined by the physical characteristics of the witnesses themselves. Hence, to evaluate the likely accuracy of witnesses’ descriptions it is worthwhile taking a description of the witness. The more similar the descriptions are to one another, the more accurate the offender description is likely to be.

 The characteristics of the witness  

Research has identified a number of witness characteristics that may also affect the accuracy of eyewitness accounts.

In general, children are poorer than adults in aspects of the encoding, storage and retrieval of information and this results in poorer memory. The ability of children to give accurate accounts, however, depends on how they are interviewed. 

The ability of older people to act as an eyewitness is not as good as that for younger or middle aged people. As we age our perceptual abilities decrease, and from 40 years onwards this decline speeds up. Additionally, older adults may have greater difficulties in storing and retrieving information. 

People with learning disabilities vary considerably in their ability to report details, and are sensitive to the ways in which questions are asked. They can, however, often provide accurate accounts of events that they have witnessed. 

Although direct assessments of the influence of mental illnesses such as depression and schizophrenia on eyewitness memory are rare, research suggests that they can have an adverse effect. 

Alcohol and drugs can have a negative effect on memory performance. This effect is related to the amount and type of the substance consumed. 

Head injuries that result in unconsciousness, even momentarily, usually lead to memory problems. Although a victim who has been knocked unconscious might not be able to recall critical information at first, memory can return to them over time. 

Interviewing eyewitnesses 

The way in which witnesses are interviewed and their responses interpreted has an important influence on the quality of the information retrieved.  

Research suggests that eyewitnesses are very good at answering questions about ‘what happened’. Hence a witnesses account of the actions of an offender can often be accurate even if poor or inaccurate details are given concerning the description of the offender.  

Witnesses tend to provide the most accurate answers to open questions. Even those groups with generally poorer memories can show high accuracy rates for open questions. As questions become more specific and closed, so accuracy rates fall. Nevertheless, ‘closed’ questions (such as ‘what colour was his shirt?’) may have to be used as a last resort to elicit information about something a witness may have omitted to describe in free recall. 

The subtle use of language in questions can influence witnesses by suggesting a particular response. The phrasing of questions which suggest a particular response, for instance leading questions such as ‘was his shirt red?’’, can decrease accuracy levels even in normal adults. 

Particular issues need to be considered when interviewing traumatised witnesses. Retrieval of information in these circumstances can be improved through remembering what was experienced emotionally during an event in order to provide retrieval cues. 

Although it is usually assumed that witness confidence is correlated with the accuracy of the information that they provide, this is not always the case. The style of questioning can influence the extent to which a witness feels confident about information. This can lead to an over - optimistic assumption about the accuracy of the information provided. Research in controlled situations, however, shows that where there has been no attempt to influence the witness, those who claim to be ‘absolutely certain’ about something tend to be correct. 

Information from identifications 

Witnesses can be good at recognizing people and items from an event even though they cannot accurately describe or recall them. Research shows a reasonable, but far from perfect, relationship between witness confidence and identification accuracy when witnesses actually make an identification at an identification parade. 

Recommendations 

1. It is important that investigators understand how to assess the reliability of the information that witnesses provide. Hence the contents of this report need to be brought to the attention of police trainers, crime investigators, crime analysts and those responsible for interviewing witnesses and victims. 

2. It is essential that those officers who interview victims and witnesses are trained in the correct interview techniques and that they employ these skills during interviews. 

3. Since the ability of a witness accurately to describe an offender will partly be determined by his or her own physical characteristics, it will help routinely to record these details. 

4. Much of the research into eyewitness accuracy has been based on laboratory studies. Whilst their findings are helpful for interpreting information from an operational setting, they are a poor substitute for the actual conditions, which victims and witnesses endure. Hence more empirical research is needed based on actual cases. In particular, the factors, which influence the accuracy of facial composites and the usefulness of these procedures to police investigations, should also be the subject of further study. 

5. Eyewitness information is important. The police should ensure that accurate records of offender descriptions are kept, and that information systems are designed to cope with the variability of descriptions.

 

To view the full paper click the URL below