"There are a bunch of studies that explain in great detail why it doesn't work, but nothing has apparently persuaded people to stop using these things," Watson said.
In a 2008 study, Watson and two other researchers surveyed nearly 400 people who were asked to compare a set of 12 composites with 12 photos and say whether the sketch matched the person. About 30 percent of the respondents thought the sketch matched its corresponding photo. But, it turns out, all of the sketches depicted the person in the matching photo.
In his coming research, Watson will insist that police departments come up with a way to calculate how effective their artists are and develop a system to decide whether the benefits of composites outweigh the possible risks.
Conlon doesn't believe composites will die. She said she plans to hone her skills in the coming year, attending more training sessions and planning to earn a certification. She hopes her work leads to more closed cases, like the man suspected of attacking the woman and teen in Chillum.
The teenager remembered the dazed look and gaunt face of the man who attacked her, Conlon recalled. Conlon pieced the memories together.
After seeing the final image, the teenage victim "started to get freaked out," Conlon said. "Her reaction when she came back into the room, that is rewarding in a strange sort of way."