Home > Development, Memory > Exonerate the Innocent!

Exonerate the Innocent!

Many innocent people are wrongly convicted of crimes every year, and many of these wrongful convictions are due to a mistaken identification during eyewitness testimony. In many criminal investigations, eyewitness identification can be a deciding factor in the case. The Innocence Project (2012) has exonerated 289 people in the U.S. based on DNA evidence. About 75% of those wrongfully imprisoned were people mistakenly identified in a line-up. (To learn more about the Innocence Project, click here.) Surprisingly, recent data have shown that approximately a third of witnesses for line-ups are children younger than 16 years old. The data also show that about a third of these children under 16 are likely to make a false identification of an innocent person as the culprit. It goes without saying that there can be very serious and severe outcomes for people as a result of false identification. For these reasons, research on eyewitness testimony has become more important and prominent in recent times. 

Because children are more likely than adults to make false identifications, researchers have been interested in why this problem occurs in children.  Eyewitnesses are shown a sequential line-up that either has the suspect or target present (TP) or a line-up with the target absent (TA). Studies contain this TA option to portray the fact that sometimes police place the focus of a crime on an innocent suspect. An example of a sequential line-up is as follows: a child is presented with the line-up members one at a time, he makes a decision on each line-up member (yes, no, or not sure) before he is shown the next person in the line-up. In a TP line-up, the child (over the age of 5) is as likely as an adult to correctly identify the suspect. However in a TA line-up, the child often mistakenly identifies an innocent suspect. This is because they either pick the person that looks most like the suspect that they had previously seen, or because they are unsure and would rather give a wrong answer than no answer because of an implicit pressure to respond and to give a positive, favorable response.

This study by Harvard and Memon (2013) included the presentation of a short film in which a young Caucasian man committed a crime. The culprit was seen in different angles in the duration of the film: full-face frontal and both profile views. The witnesses (children ages 5-11) were presented with a line-up 1 or 2 days after having seen the film at their school. Half of the line-ups were TP and half were TA. The witnesses were presented with a 15-second clip of the person looking straight, left, right, and back to full face. The “mystery man” line-up was created with a computer program so that it could move as well. See Figure 1 below.

Mystery Man

As expected, with the addition of the “mystery man”, false identifications were reduced from 75% to 40%. This “mystery man” provides the child with an option of giving a positive response and not falsely identifying a person just because of the pressure put on the witness to make a selection.

These results can be largely beneficial in real criminal investigations. With no adverse effect on correct identifications, there seems to be no reason to not always include the “mystery man” in a line-up. Because the potential punishments and consequences can be so severe for people wrongly imprisoned, this simple implementation of a “mystery man” into line-ups can make a world of difference. Providing this extra option of a “mystery man” reduces chances of falsely identifying an innocent person, as well as giving witnesses who are not completely confident in their suspicions a way of not falsely identifying someone. This small option created by a computer program could change the course of a person’s life, in that an innocent person might not be wrongly identified as a criminal, and so I feel that all police forces should be required to include the “mystery man” in future line-ups that include children as eyewitnesses.

To read the original article, click here.

To read more blogs on eyewitness testimony, click any of the following links:

How to Steal an Exit Sign

Innocent Criminal

Remembering the Deceitful 

Faulty Eyewitness Testimony


Havard, C., & Memon, A. (2013). The mystery man can help reduce false identification for child witnesses: Evidence from video line-ups. Applied Cognitive Psychology, 27(1), 50-59. doi:10.1002/acp.2870


  1. November 28th, 2013 at 20:14 | #1

    Although the study is about children and how they falsely identify criminal, I think it would be really interesting to do the study on adults and compare them to these results. There are so many problems with convicting people in lineups due to false memory, transience, bias and lack of attention so I think it would be great if we could give every witness a better opportunity to convict the right person. If another study found that this online program was even slightly helpful I think it would be a great starting point for future research and improvement in this area of the incriminating system.

  2. December 1st, 2013 at 20:24 | #2

    This seems like such a simple and elegant tool for working with child witnesses. We’ve discussed in class children’s desire to please, and how in many experimental settings this can lead to “false alarm” responses instead of correct rejections. It makes a lot of sense, then, that this would carry over into real-world situations like witness identification. I agree with Kimberly – I definitely think it’s something worth exploring with adults (maybe particularly older adults), because I would imagine that witnesses, no matter what their age, would always feel a certain amount of pressure to identify someone in the line-up.

  3. December 2nd, 2013 at 14:53 | #3

    This is an interesting read. I am also interested in seeing how adults would compare to children. Since children usually give incorrect responses to make a good impression, it would be interesting if adults also have a desire to please. As discussed in class, there are so many things that affect witness testimony such as emotions, bias, and transience. Also, slight changes in a person’s description can either lead to their prosecution or release. For example, if the witness did not have a good view of the person, they would usually look for or remember something distinctive about that person. If you watch the TV show, “Person of Interest” the man in the suit is a good example. Because no-one knew his face, In this show, police went around looking for a man in a suit regardless of the fact that in a day, more than 10 people wear suits. I therefore think a system, like the “mystery man” that prevents such generalizations would definitely be helpful to a lot of people.

  4. December 2nd, 2013 at 20:02 | #4

    Having taken Martha’s child cognitive development seminar focusing on the reliability of children’s eye-witness testimony last fall, this was a very interesting read! After taking PS232, however, I can now view this issue from a new perspective that includes the many facets of memory and how it can fail us! As has been touched upon in previous comments, our memory can fall to the seven sins of memory, such as transience, which is the gradual loss of memory over time, or suggestibility, which is the addition of false information into a memory. Therefore, the statistics about the susceptibility of false recall for children doesn’t exactly surprise me. As was discussed in class, it is also surprisingly easy to implant a false memory! For example, Loftus & Picknell (1995) were able to make 25% of participants create an event that never happened! Therefore, providing children with the option and comfort of a “positive response” takes away the pressure of wanting to believe that the culprit is in the line-up, just as the participants in the Loftus & Picknell (1995) study wanted to believe that the false event really happened. I believe that this is a great advancement in the field of eye-witness testimony!

  5. December 3rd, 2013 at 20:57 | #5

    This was a very interesting and important study that could help improve procedures used in our judicial system. Children’s eyewitness testimony has many other issues in addition to the forced-choice element, which all need to be addressed in order to yield the most accurate responses possible. In Dan’s post, “The Worst Lies Involve Raccoons,” he talked about source monitoring causing problems for adult eyewitness testimony. This past semester, I also took the cognitive development seminar, and we talked about this phenomenon as an obstacle in the context of children’s testimony as well. We looked at a study by Poole and Lindsay (2002), who tested a source monitoring training procedure that helped children identify memory sources and reduced false reports in 7- and 8- year-olds. If more studies like those by Poole and Lindsay and Karageorge and Zajac are conducted, we can drastically improve the eyewitness testimony of children through looking at many potential sources of error, one by one. This is very promising for a field plagued by inaccuracy, and small improvements on the way children are questioned could add up to make great strides.

  6. Tara Nguyen
    March 19th, 2014 at 15:43 | #6

    It’s relieving to know that there’s at least a way to reduce false conviction of innocent citizens. That said, I have a few questions regarding the line-ups.
    – What would happened if children (or witnesses in general) viewed all suspects at the same time instead of one suspect at a time? In this case, for TA trials, would witnesses make comparisons between suspects to decide which one was the most familiar?
    – What if race were to be taken into account? As Lynna mentioned in her post, “How to Steal an Exit Sign,” research by Horry and Wright (2008) showed that people are worse at cross-race recall than at same-race recall. This can be explained by the own-race bias, which says that people recognize faces of their own race better than they do faces of other races. However, because processing of faces improves with age and expertise (e.g., Rhodes et al., 1989; Valentine, 1991; as cited by Harrison and Hole, 2009), children may not be good at face recall regardless of race. Thus, I wonder if children would be able correctly identify a suspect if the suspect were of a different race from the children.
    In line with the own-race bias, the own-age bias may also affect children’s identification of suspects. Because the suspects are (supposedly) adults, children may have difficulty making the right decision when it comes to face recognition.

You must be logged in to post a comment.