Eyewitness Identification

Research: Estimator and System Variables

Although eyewitness identification has long been a staple of evidence used to convict criminal suspects, psychological scientists have conducted hundreds of staged-crime experiments since the late 1970s that show alarming rates of mistaken identification. Courtroom challenges to the reliability of eyewitness identification evidence by psychological scientists have become increasingly common in the United States. In the 1990s, the development and use of forensic deoxyribonucleic acid (DNA) managed to corroborate the general claims that psychological scientists had made about the reliability of eyewitness identification evidence. Individuals who had been convicted of serious crimes later had their innocence proved with DNA testing; these were primarily cases of mistaken eyewitness identification.

Eyewitness identification researchers have discovered scores of variables that affect the accuracy of eyewitness identification. These include characteristics of the witnesses, expectations before witnessing, witnessing conditions, events occurring after the witnessed event, and the methods used to obtain the identification. A useful categorization of these variables separates them into estimator variables and system variables. Estimator variables affect the accuracy of eyewitness identification but cannot be controlled in actual cases. Examples of estimator variables include characteristics of the witnesses, lighting conditions at the time of witnessing, and so on. System variables also affect the accuracy of eyewitness identification but have the additional characteristic of being controllable in actual cases. Examples of system variables include instructions to witnesses before viewing a lineup, the structure of the lineup, and so on. This dichotomy has proved useful because it manages to inform the justice system about ways to improve the accuracy of eyewitness identification evidence (the system-variable focus).

The most researched estimator variables include stress; higher levels interfere with forming a clear memory of faces. Another is the race or ethnicity of the witness and culprit. Within-race or within-ethnicity identification of faces is more reliable than are crossrace or cross-ethnicity identifications. Finally, an eyewitness's self-rated confidence in an identification that was made has a modest, but statistically reliable, positive correlation with eyewitness identification accuracy.

The most researched system variables include prelineup instructions to witnesses. Witnesses who are warned that the actual culprit might not be in the lineup make fewer mistaken identifications. Another is the composition of the lineup; using fillers who do not fit the description of the culprit leads to increased risk of mistaken identification. Finally, the method of presentation is important. Showing all lineup members at once, known as a simultaneous lineup, encourages more guessing than does showing each lineup member individually, which is known as a sequential lineup.

...