Psychologists Gary Wells and Brian Cutler helped design a procedure in 2003 for witnesses to identify crime suspects. Police departments across North Carolina embraced the procedure. The Durham Police Department adopted it almost word- for-word in February.

The conduct of the Durham police in the Duke lacrosse case, however, is a case study in violating the new policy, the psychologists said. And as a result, police have injected doubt into a woman's selection of three lacrosse players whom she accused of rape.

Police violated two fundamental rules for running an identification procedure, said Wells, a professor at Iowa State University, and Cutler, a professor at UNC-Charlotte.

First, the psychologists said, police did not have an independent investigator administer the process. Second, they neglected to include photos of nonsuspects, known as fillers.

The procedures used can yield only uncertain or misleading results, Wells said, and that's bad for everyone.

If the woman was raped, Wells said, the botched lineups undermine the prosecution and the search for justice.

"And if she wasn't sexually assaulted, or was assaulted by someone else, [the players are] in a position of guilty until proven innocent," Wells said. "It really shifts the burden to the person identified to prove it wasn't them. That is a profoundly difficult and very unfair situation."

Compounding the flawed procedures are the accuser's attempts to pick out her alleged assailants. She selected 20 players when she viewed their photographs in March and April. Every choice contained flaws or contradictions, according to an analysis of police records turned over to defense lawyers by District Attorney Mike Nifong.

The woman recognized 15 players at one viewing but didn't recognize them at another.

She picked out only one player with certainty at both the March and April viewings. He, however, was in Raleigh, not at the lacrosse party.

She wrongly identified the player who made a rude comment about a broomstick.

The woman, a dancer for an escort service, picked out four assailants in April after telling police that three men gang-raped her March 14 in a bathroom at a lacrosse team party.

There are problems with her identification of the three men who have been indicted. She said David Evans had a moustache; his lawyers say he never had one. She identified Collin Finnerty as an assailant; Finnerty did not match any of the initial descriptions she gave police. In March, she was 70 percent sure that she recognized Reade Seligmann but couldn't recall where she saw him at the party. In April, she was 100 percent certain he had orally raped her.

Arranging lineups

Cutler, the UNC-Charlotte psychologist, said a witness may identify a suspect for the correct reason: The suspect is the perpetrator.

But, he said, there can be other reasons: The witness could be guessing; the witness could be lying; the witness could be influenced by police behavior to pick a suspect; the witness could be influenced by repeated questioning or by having seen a picture on television or in the newspaper.

"A well-conducted test will maximize the possibility that the reason for identification is recognition and will rule out the other nuisance or extraneous results," Cutler said.

In the lacrosse case, Durham police conducted several ID sessions with the accuser. On March 16, she viewed 24 photos of lacrosse players printed from the goduke.com Web site. On March 21, she viewed 12 more.

She recognized five of the 36 players, but police records of the March lineups are unclear whether she identified the five players as assailants or merely as partygoers. According to forms filled out at the time by Investigator Richard Clayton, he showed the accuser a series of photographs and asked whether the person had sexually assaulted her.

Next page >

Get $150+ in coupons in every Sunday N&O. Click here for convenient home delivery.