How Verbal Descriptions Affect Identification of Criminals

New research extends studies of how talking about faces affects identification.

Posted Mar 13, 2018

IMDB
Source: IMDB

If you have ever watched a police procedural show on TV, you’re familiar with the typical pattern of investigation of a crime.  A crime is committed and then the police are called.  They interview witnesses and go out looking for suspects.  A suspect is found and shown to the witnesses using a lineup.  If the suspect is identified from the lineup, that is a piece of evidence used against them at the trial.

Lots of research in psychology has shown that it is actually hard to construct good lineups, because many aspects of the way lineups are constructed can bias the judgments of witnesses. 

One fascinating finding that causes potential problems for eyewitness memory is the verbal overshadowing effect.  In studies of this effect, participants witness a crime and then later describe the perpetrator.  Compared to people who do not give a description, those who described the perpetrator are less accurate at identifying the criminal later from an array of pictures.

A paper by Brent Wilson, Travis Seale-Carlisle, and Laura Mickes in the January, 2018 issue of the Journal of Experimental Psychology: General  explored this phenomenon in more detail to understand why it happens and to provide more detail about what is going on.

Across several studies, they varied both the timing of the verbal description participants gave as well as the construction of the lineup.  In each study, participants first watched a video depicting a crime.  Then, there was a 25-minute delay during which participants did puzzles or played games.  Some participants described the perpetrator immediately after watching the video.  Some participants described the perpetrator after a 20-minute delay.  Some participants never described the perpetrator.

In some studies, a traditional lineup was given in which participants saw 6 pictures.  Some lineups had the actual perpetrator in it.  Some did not.  Participants had to identify the perpetrator with the option to say that the perpetrator was not in the lineup.  In other studies, a “showup” was used in which participants saw just one picture that was either of the perpetrator or of another person and they had to identify whether the picture was the perpetrator.  After making their judgment, participants also rated how confident they were that their judgment was correct.

The type of lineup used didn’t matter. The results were the same for the lineup or the showup.

When participants described the perpetrator immediately after seeing the video, describing the person had no impact on their accuracy compared to the control condition.  In addition, the more confident people were that their judgment was correct, the more accurate they were.

When participants described the perpetrator after a 20-minute delay, those who gave a description were significantly less accurate at identifying the perpetrator than those who did not describe the perpetrator.  So, there is a verbal overshadowing effect.  However, the participants who were most confident in their judgments were equally good at identifying the perpetrator regardless of whether they gave a verbal description.

Why do descriptions only affect eyewitness identification performance after a delay?

The researchers looked at the features people used to describe the perpetrators.  Compared to the descriptions given after a delay, the descriptions given immediately were more likely to include distinctive characteristics of the perpetrator that would distinguish them from the other people in the lineup.  After a delay, some of the details about the faces may have faded from memory enough that the delayed descriptions of the faces were more generic. 

As another test of this possibility, another group of participants were shown only a description of the perpetrator given by another participant and were asked to identify the perpetrator from the lineup.  Participants were better able to identify the perpetrator when given a description generated by someone immediately after seeing the video than when given a description generated by someone after a delay.

There are several interesting aspects of this work.

First, if witnesses are going to give a description of people involved in a crime, then they should do so immediately.  Second, although confidence in memories is not always related to accuracy, it appears that confidence matters in eyewitness situations.  Third, investigators should look carefully at whether the descriptions witnesses give contain differentiating information about perpetrators.  If not, it may be that those witnesses will have trouble with lineups.

References

Wilson, B.M., Seale-Carlisle, T.M., & Mickes, L. (2018). The effects of verbal descriptions on performance in lineups and showups. Journal of Experimental Psychology: General, 147(1), 113-124