Is an eyewitness’s confidence the key to their reliability?

When police ask an eyewitness to identify a suspect, it is usually done through either a lineup or a photo array. In this situation, the culprit may actually be among the choices. However, it’s also possible that the police have the wrong person. Therefore, identifying a suspect should always be about affirmatively recognizing them, not just choosing the person most similar to the suspect.

Confidence game

Police, prosecutors, courts and juries, along with other experts like psychologists, have traditionally relied on the degree of confidence the witness expresses. The more confident the witness, the more certain they feel that the witness has actually recognized the suspect.

But, over the course of decades, groups like the Innocence Project have shown that eyewitness identifications are often wrong — even when the witness indicated strong confidence in his or her choice. Many people have been exonerated by DNA or other evidence despite an eyewitness having confidently identified them as the perpetrators.

Do witnesses’ expressions of strong confidence have any relationship with whether they’re objectively correct in their identification? For many years, researchers believed there was only a weak relationship between witness confidence and objective accuracy. However, in the 2000s, a new method of analysis seemed to demonstrate that high confidence provides a reasonable — if imperfect — indication of accuracy.

Does that mean that a highly confident witness is almost certainly correct? The new analysis does seem to indicate that — as long as the lineup or photo array was administered in “pristine conditions.”

What are pristine conditions in a lineup or photo array?

Not all lineups and photo arrays are done fairly, even when the police aren’t trying to cheat. We can all imagine unfair conditions. For example, it wouldn’t be fair — yet it might lead to a highly confident witness — if the perpetrator was red haired and all the other lineup members had black hair.

There are more subtle ways police can affect the witness’s choice, too. Simply knowing which of the lineup members is suspected can lead officers to subtly push the witness in the “right direction.” This can lead to an innocent person being identified as the culprit.

So, psychologists and other experts have come up with rules for a fair identification process, which are called “pristine conditions”:

  • The lineup must contain only one suspect
  • The other members of the lineup must be completely excluded from suspicion
  • The other members of the lineup must be similar in appearance to the suspect
  • The administrator must warn the eyewitness that the culprit might not be among the choices
  • The administrator must not know which choice is the suspect
  • The witness’s confidence must be measured right away, before any communication occurs between the witness and the administrator

When these pristine conditions are in place, we can be reasonably sure that the police and conditions do not lead the witness to choose a particular person that the police suspect. Unfortunately, few law enforcement jurisdictions actually have policies promoting the use of pristine conditions. Lawmakers simply haven’t implemented them or committed to monitoring them.

Furthermore, in real-world conditions it can be difficult to set up a lineup, or even a photo array, that contains a group of similar-looking people. Even if you have five redheads available, they might not be the same type, for example.

Therefore, there is much more research to be done before we can determine if a witness’s confidence in an identification actually indicates that identification is likely to be accurate. Until we know more, we should always question how witness identifications are achieved.

Skip to content