They may work on ”Law & Order” or in the classic film “The Usual Suspects,” but researchers say that the traditional police lineup could use an overhaul.

“Mistaken eyewitness identification is the primary cause of the conviction of innocent people,” said Dr. Gary Wells of Iowa State University.  “If you look at, for example, DNA exoneration cases where DNA came along and was used to test claims of innocence…three out of every four of those exonerations are cases of mistaken identification.”

Getty Images

Wells, an eyewitness identification expert and the lead researcher behind a new study by the American Judicature Society, says that police forces should update the way they conduct lineups. 

“One of the things we observed over and over again is that witnesses tend to compare one person to another, decide who looks most like the perpetrator and then their propensity is to pick that person.  That’s OK if the real perpetrator is in there, but if the real perpetrator is not in there, there’s still someone that looks more like the perpetrator than the others,” Wells said.

The standard lineup seen on TV and in movies has witnesses looking at groups of people either standing in front of them or in a grouping of photos.   Researchers call this a simultaneous lineup.  Wells and his team said their research shows that a sequential lineup is more effective.

In a sequential lineup, a witness looks at each person individually and says whether he or she is the suspect before moving onto the next person.  Wells said that when a witness looks at each person individually in a lineup, they are less likely to mistakenly pick one of the “filler” people added to fill out the lineup. 

“The sequential procedure reduced these types of mistaken identifications without reducing the frequency of which witnesses pick the suspects,” he said.

Wells and his team worked with the police departments from Austin, Texas, Tucson, Ariz., San Diego, Calif., and Charlotte, N.C., testing the two methods of lineups. Each detective conducting the different lineups was also blind as to which person in the lineup was the suspect and which were just fillers. Researchers say that this also helps more effective identification.

The statistical analysis of the study revealed that the two procedures yielded similar numbers in identifying the suspect.  However, the sequential method yielded fewer misidentifications. Only 12.2 percent of the time did the witness misidentify the suspect. The more traditional simultaneous method yielded the mistaken identification of a filler person 18.1 percent of the time.

Along with producing fewer mistaken identifications, the sequential method also showed that witnesses were more confident, producing fewer “not sure” responses.

Wells said that some police departments have already changed the way they do lineups over the last few years.

“A number of jurisdictions- New Jersey, North Carolina, places like Boston, Denver, Dallas…but there are still so many jurisdictions that sort of have resisted this switch,” Wells said.