Hundreds of individuals are chosen from lineups every day in North America and Europe (Smith, Lindsay, Pryke, & Dysart, 2001). According to The Innocence Project (2008) “Eyewitness misidentification is the single greatest cause of wrongful convictions nationwide, playing a role in more than 75% of convictions overturned through DNA testing.” Still, the criminal justice system profoundly relies on eyewitness identification and testimony for investigating and prosecuting crimes (Wells & Olson, 2003).
I will now provide a brief overview of factors, including both estimator and system variables, which influence the accuracy of eyewitness accounts. First, I will look at estimator variables and their effect on eyewitness accounts. Second, I will examine system variables, how they affect eyewitness accounts, and how the criminal justice system can control them. And finally, based on my research I will offer suggestions to increase the accuracy of eyewitness accounts.
Estimator variables are variables that influence eyewitness accounts and identification that are out of the hands of the criminal justice system. Estimator variables can potentially distinguish accurate from inaccurate eyewitness identifications (Wells & Olson, 2003). Estimator variables include both characteristics of the witness, such as age and race, and characteristics of the event, such as lighting conditions and time of day (Smith, Lindsay, Pryke, & Dysart, 2001).
Studies have shown that witness characteristics can have a profound effect on eyewitness recall and identification (e.g. Wright & Stroud, 2002). Some researchers have argued that age is an extremely important eyewitness factor (e.g. Kwong See, Hoffman & Wood, 2001), while others focus on factors that can be changed, such as mood (e.g. Forgas, Laham & Vargas, 2004) and drug use (e.g. Yuille, Tollestrup, Marxen, Porter, & Herve, 1998).
Many factors affect the ability of an eyewitness to recall the event at a later time. Some studies have examined event characteristics directly related to the culprit, while others have looked at the impact of event characteristics with no relation to the culprit (Wells & Olson, 2003).
Regardless of what specific factor is studied, researchers can agree that witness and event characteristics do indeed influence eyewitness accounts.
Characteristics of the Witness
Eyewitness identification performance has been consistently linked with the age of the eyewitness. A well-known elderly stereotype is that with old age comes poor memory. Kwong See, Hoffman, and Wood (2001) studied this stereotype in hopes of learning more about age and eyewitness memory and found that while older witnesses are described as more honest, younger witnesses are described as more competent. Similarly, Pozzulo and Lindsay (1998) studied eyewitness identifications errors in relation to age and found that when the lineup contains the actual culprit young children and elderly perform well, but when the lineup does not contain the culprit there is a higher rate of mistaken identifications.
The race of the witness has also been studied at length. The Cross Race Effect (CRE), also known as the Own-Race Bias or the Other-Race Effect (ORE) is a phenomenon from which people are better at recognizing faces of their own race rather than faces of other races (Michel, Rossion, Han, Chung & Caldara, 2006). Meissner and Brigham (2001) recently constructed a meta-analysis that follows the cross race effect for more than twenty-five years.
A number of studies have suggested that positive moods may lead to effortless systematic processing while negative moods facilitate more alert systematic processing (Schwartz & Bless, 1991, cited in Forgas, Laham, & Vargas, 2004). Perhaps the explanation for this is that in attempt to preserve positive moods, cognitive effort is avoided; to repair negative moods, cognitive effort is increased (Clark & Isen, 1982). Forgas, Laham, and Vargas (2004), studied the effect of temperament on eyewitness memory and concluded that those in negative moods experienced more false alarms.
Some research has been conducted to distinguish how the effects of drugs and alcohol can affect eyewitness recall. Yuille et al. (1998) found that marijuana only has a temporary negative effect on the amount of information witnesses can remember, and overall, has little influence on recall accuracy or recognition. Alcohol, on the other hand, has permanent consequences for memory. Yuille and Tollestrup (1990) found that witnesses do not show improvement in eyewitness accuracy between their intoxicated recollection immediately after the event and their sober attempt.
Characteristics of the Event
As to be expected, the most important event characteristic is the offender. Whether they are wearing a disguise, if they are carrying a weapon, and how much time they are in view of the witnesses are important in relation to eyewitness accounts. Other important event characteristics include lighting conditions and the timing of knowledge that the observer is indeed witnessing a crime (Wells & Olson, 2003).
Studies have shown that distinctive faces and faces that are either highly attractive or highly unattractive are more likely to be accurately recognized. Even the simplest disguises can impair accurate eyewitness identification. If an offender is wearing a hat, sunglasses, or a sweatshirt with a hood, they significantly increase the chances of misidentification. Clearly, if an offender is wearing a full facial mask or disguise, the chances of them being identified later decline. In the case that the offender is wearing such a disguise, but still displays tattoos or body piercings, accurate identification chances rise (Wells & Olson, 2003).
Weapons have an extreme impact on eyewitness accounts. Steblay (1992) reported that weapon presence significantly reduces the probability that the witness will be able to later identify the offender. The weapon focus effect is the observation that if there is a weapon at the scene, witnesses are likely to focus on the weapon instead of the offender’s face. As result of weapon focus, witness recall of offender appearance is less clear (Brain, 2002). The presence of threatening stimuli can lead to fear and emotional stress, which can influence eyewitness memory (Wells & Olson, 2003).
The amount of time an offender is in view of the witness is important when it comes to identifying the offender at a later time. Perhaps what is more important is how much attention the witness directs toward the offender. If the offender is only in view for a minute during which time the witness is not paying attention the chances of later identification is much worse than if the offender is in view for a couple minutes and the witness is focusing all of their attention on the offender. The amount of attention is directly influenced by how long it takes the witness to realize they are witnessing a crime. Furthermore, lighting conditions and weather can also influence what a witness will be able to observe about an event or culprit (Wells & Olson, 2003).
Context also immensely impacts eyewitness accounts. Ainsworth (2002) explained that information is observed depending on the way in which we perceive it. For example, a man wearing all black and a ski mask is perceived much differently at a ski resort than at a bank (Leinfelt, 2004).
During the Washington D.C. sniper case the public was bombarded by the media begging anyone with information about a white van or number of shooters to come forward and as a result many more people came forward than anyone expected. This could be because 1) the media manipulated their memory, 2) visual hindsight bias, or 3) they were filling in the gap. Highly publicized crimes generally have higher misidentification rates because providing the public with information may cause good eyewitness accounts to turn bad. According to Leinfelt (2004), memories can be manipulated by information the witness is exposed to shortly after an event occurs. Hindsight bias, or the knew-it-all-along effect is the tendency to claim to know the outcome or at least have a higher estimate of it after the fact (Harley, Carlsen, & Loftus, 2004). In simple terms, it is the “I told you so” or “I knew this was going to happen” response. Filling in the gap is a similar task; an example of this is when someone fills in the missing parts of a story without actually experiencing it. In a study by Gerrie, Belcher, and Garry (2006), forty-four subjects watched one of two versions of a video showing a woman making a peanut butter and jelly sandwich, each video was missing three different steps of the process (1, 3, 5; 2, 4, 6), and forty out of the forty-four subjects falsely recognized at least one of the missing clips.
System variables are variables that influence eyewitness identification that are able to be controlled by the criminal justice system. System variables can potentially prevent inaccurate identifications (Wells & Olson, 2003). System variables generally tend to include lineup characteristics, administrator behavior, and instructions (Smith, Lindsay, Pryke, & Dysart, 2001).
Much research has been done in relation to lineup characteristics and eyewitness accuracy (e.g. Wells, Rydell, & Seelau, 1993). There has also been a noteworthy amount of research done on administrator behavior (Wells & Olson, 2003) and instructions (Steblay, 1997). Regardless of what specific factor is studied, researchers can agree that all of the above system variables do indeed influence eyewitness accounts.
Wells and Olson (2003) describe a culprit-absent lineup as “one in which an innocent suspect is embedded among fillers,” while a culprit-present lineup is “one in which a guilty suspect is embedded among fillers” (p. 279). In any situation, a lineup being shown to a witness may or may not include the actual culprit. Culprit-absent lineups most frequently occur when police unknowingly focus their investigation on an innocent person. As could be expected, culprit-absent lineups cause the most problems for eyewitnesses, perhaps because one usually expects the culprit to be present in a lineup when they see it (Wells & Olson). In a study by Wells, Rydell, and Seelau (1993), participants were shown a five-person culprit-absent lineup in which only 32% made no identification, while the other 68% mistakenly identified.
In relation to lineup content, non-suspect fillers are chosen to be present in a lineup. Fillers are chosen to fit both the verbal and the visual description given by the witnesses of the offender. Ideally, lineup fillers are chosen so that an innocent person is not mistakenly identified for standing out and so that a guilty person is not overlooked just for blending in. While selecting fillers to fit the overall description of the suspect can decrease misidentification rates in culprit-absent lineups, they can unfortunately increase misidentification rates in culprit-present lineups (Wells & Olson, 2003).
There are three main types of lineup presentations: Simultaneous, sequential, and blank. Simultaneous lineups present all lineup members to the eyewitness at once and these are used most frequently. Simultaneous lineups are also the most erroneous. Sequential lineups present the eyewitness with one lineup member at a time, with the witness knowing that there are several lineup members to be shown. Sequential lineups can prevent false identifications because witnesses are not as likely to choose the lineup member that looks most like the culprit compared to other lineup members since they will see all lineup members separately. Blank lineups contain only fillers, are shown first with the expectation that it is the only lineup that will be shown, and if the witness identifies from the blank lineup they will be discarded (Wells & Olson, 2003).
More often than not, the person administering a lineup is also the case detective who obviously knows which one of the lineup members is the actual suspect. Lineup administrators could communicate, either verbally or non-verbally, with the witness to suggest that one lineup member is the culprit over the others. Lineup administrators can also increase witness levels of uncertainty after identification by making post-identification suggestions. Furthermore, if an administrator has any sort of extreme mood, whether it be positive or negative, the witness could be intimidated by them and make a rushed identification as a result (Wells & Olson, 2003).
Pre-lineup instructions have a substantial impact on eyewitness identification. Steblay (1997) found that misidentification rates are reduced (by 41% in her study) when witnesses are warned that the culprit may or may not be in the lineup, instead of not receiving any instructions at all. On the other hand, giving leading instructions or making suggestions will greatly increase misidentification rates. However, double-blind testing, assigning a lineup administrator who does not know the case or who the actual suspect in the lineup is, can help prevent this problem (Wells & Olson, 2003).
In my research I came across a few suggestions to help improve accuracy of eyewitness identification:
Obtain the witnesses statement as close to the event as possible
Use open-ended questions when interviewing the witness
Have the witness tell the story from beginning to end and from end to beginning, in attempt to prevent missing any important details
Record all questions, both asked and answered, so that there is record and the ability to refer back if necessary
Avoid interrupting the witness while he or she is giving a statement, being interviewed, seeing a lineup, or testifying.
Avoid repeating questions during the interview process
Include blank lineups, as they are an easy way to discard unreliable witnesses
Instruct witnesses that the culprit may or may not be in the lineup so they do not feel pressured to make an identification
Use a sequential lineup presentation to avoid witnesses comparing the lineup members
Use double-blind testing, so that the person administering the lineup cannot influence the witness
Use caution when/if providing feedback about correctness of identification
There are two types of variables that influence eyewitness accuracy: estimator variables and system variables. Estimator variables are variables that influence eyewitness accounts and identification that are out of the hands of the criminal justice system. Estimator variables can potentially distinguish accurate from inaccurate eyewitness identifications (Wells & Olson, 2003). Estimator variables include both characteristics of the witness, such as age and race, and characteristics of the event, such as lighting conditions and time of day (Smith, Lindsay, Pryke, & Dysart, 2001). System variables are variables that influence eyewitness identification that are able to be controlled by the criminal justice system. System variables can potentially prevent inaccurate identifications (Wells & Olson, 2003). System variables generally tend to include lineup characteristics, administrator behavior, and instructions (Smith, Lindsay, Pryke, & Dysart, 2001).
Eyewitness identification performance has been consistently linked with the age of the eyewitness. The race of the witness has also been studied at length. Studies have shown that people in negative moods facilitate more alert systematic processing and therefore can better recall information than those in positive moods. Some research has been conducted to distinguish how the effects of drugs and alcohol can affect eyewitness recall. As to be expected, the most important event characteristic is the offender. Whether they are wearing a disguise, if they are carrying a weapon, and how much time they are in view of the witnesses are important in relation to eyewitness accounts. The amount of time an offender is in view of the witness is important when it comes to identifying the offender at a later time. Context also immensely impacts eyewitness accounts, as does the media and visual hindsight bias.
Culprit-absent lineups create the most difficulty for eyewitnesses. Simultaneous lineups are used most frequently and are also the most erroneous. Lineup administrators sometimes communicate, either verbally or non-verbally, with the witness to suggest that one lineup member is the culprit over the others. Pre-lineup instructions have a substantial impact on eyewitness identification.
1. Ainsworth, P. B. (2002). Psychology and Policing. Devon, UK: Willan Publishing.
2. Brain, C. (2002). Advanced psychology: Applications, issues, and perspectives. Cheltenham, UK: Nelson Thornes.
3. Clark, M. S., & Isen, A. M. (1982). Towards understanding the relationship between feeling states and social behavior. In A. H. Hastorf & A. M. Isen (Eds.), Cognitive social psychology (pp.73-108). New York: Elsevier-North Holland.
4. Forgas, J. P., Laham, S. M., & Vargas, P. T. (2005). Mood effects on eyewitness memory: Affective influences on susceptibility to misinformation. Journal of Experimental Social Psychology, 41, 574-588.
5. Gerrie, M. P., Belcher, L. E., & Garry, M. (2006). ‘Mind the gap’: False memories for missing aspects of an event. Applied Cognitive Psychology, 20(5), 689-696.
6. Harley, E. M., Carlsen, K. A., & Loftus, G. R. (2004). The “saw-it-all-along” effect: Demonstrations of visual hindsight bias. Journal of Experimental Psychology: Learning, Memory, and Cognition, 30(5), 960-968.
7. The Innocence Project. (2008). Innocence Project. Retrieved November 16, 2008 from http://www.innocenceproject.org/
8. Kwong See, S. T., Hoffman, H. G., & Wood, T. L. (2001). Perceptions of an old female eyewitness: Is the older eyewitness believable? Psychology and Aging, 16(2), 346-350.
9. Leinfelt, F. H. (2004). Descriptive eyewitness testimony: The influence of emotionality, racial identification, question style, and selective perception. Criminal Justice Review, 29(2), 317-340.
10. Meissner C. A., Brigham, J. C. (2001). Thirty years of investigating the other-race bias in memory for faces. Psychology, Public Policy, and Law, 7(1), 3-35.
11. Michel, C., Rossion, B., Han, J., Chung, C., & Caldara, R. (2006, July). Holistic processing is a finely tuned for faces of one’s own race. Psychological Science, 17(7), 608-615.
12. Pozzulo, J. D., & Lindsay, R. C. L. (1998). Identification accuracy of children versus adults: A meta-analysis. Law and Human Behavior, 22, 549-570.
13. Steblay, N. M. (1992). A meta-analytic review of the weapon focus effect. Law and Human Behavior, 16, 413-424.
14. Steblay, N. M. (1997). Social influence in eyewitness recall: A meta-analytic review of lineup instruction effects. Law and Human Behavior, 21 283-298.
15. Smith, S. M., Lindsay, R. C. L., Pryke, S., & Dysart, J. E. (2001). Postdictors of eyewitness errors: Can false identifications be diagnosed in the cross-race situation? Psychology, Public Policy, and Law, 7, 153-169.
16. Wells, G. L., & Olsen, E. A. (2003). Eyewitness testimony. Annual Review of Psychology, 54(1), 277-295.
17. Wells, G. L., Rydell S. M., Seelau, E. P. (1993). On the selection of distracters for eyewitness lineups. Journal of Applied Psychology, 78, 835-844.
18. Wright, D. B., & Stroud, J. N. (2002). Age differences in lineup identification accuracy: People are better with their own age. Law and Human Behavior, 26(6), 641-654.
19. Yuille, J. C., Tollestrup, P. A. (1990). Some effects of alcohol on eyewitness memory. Journal of Applied Psychology, 75, 268-273.
20. Yuille, J. C., Tollestrup, P. A., Marxsen, D., Porter, S., & Herve, H. F. M. (1998). An exploration on the effects of marijuana on eyewitness memory. International Journal of Law and Psychiatry, 21(1), 117-128.