The rates of misidentification produced by children—that have exceeded 70% in some studies—pose challenges to enhancing witness credibility.

In searching for a technique that would sustain correct identification when the perpetrator is present in a lineup, but would decrease false identifications when he is not present, Pozzulo and Lindsay hit paydirt. The participants in their study were children, ages 10 through 14, and adults. All watched a videotape showing a male character (named Mike) discussing safety. After viewing the tape the study participants were given unrelated tasks, and then about a half-hour later were shown a six-person lineup and asked to identify Mike. Some saw a lineup with Mike present, while others saw a lineup in which Mike did not appear.

When Mike was not in the lineup, and the lineup members were viewed simultaneously, the children made more false identifications than the adults (46% versus 13%). However, if children were given "elimination" instructions, their rate of false identification dropped dramatically. Several elimination procedures were tested, including those that first asked witnesses to select the lineup member who looked most like Mike, or to eliminate lineup members who look least like Mike. One of the best procedures for the children was the "slow-modified elimination" condition, in which children were told:

"Now I'm going to show you some pictures. Mike's picture might be here or Mike's picture might not be here. To start, I would like you to think back to what Mike looks like. Please look at the pictures and pick out the person who looks the least like Mike." Once a photo was chosen, it was removed. Then the children were told to pick out the person who looks the least like Mike from the remaining photos. This process was continued until only one photo remained. At this point the children were cautioned about the possibility of wrong people being arrested, and then they were asked whether they thought the remaining photo was a picture of Mike or not.

The slow-modified elimination procedure reduced the false identification rate to 27%—quite a drop from the 46% rate with the simultaneous procedure. The investigator also has the option of skipping the last step (Is the remaining photo Mike or not?) if the suspect has already been eliminated. This prevents the child from making a false identification, and can preserve the child's credibility for other aspects of the experience. The authors should be congratulated on their efforts to find ways not of complaining about the poor performance of children, but of helping them to perform more effectively.

Pozzuio, J.D. & Lindsay, R.C.L. (1999). Elimination lineups: an improved identification procedure for child eyewitnesses. Journal of Applied Psychology, vol. 84, no.2,167-176.