This article was published on page 1 in the New York Times on July 21, 2001

New Way to Insure Eyewitnesses Can ID The Right Bad Guy

By GINA KOLATA and IVER PETERSON

Prompted by new insights into the psychology of eyewitnesses to

crimes, New Jersey is changing the way it uses witnesses to

identify suspects.

Starting in October, the state will become the first in the nation to give up the familiar books of mug shots and to adopt a simple

new technique called a sequential photo lineup, said John J. Farmer

Jr., New Jersey's attorney general. Sequential viewing of

photographs has been shown to cut down on the number of false

identifications by eyewitnesses without reducing the number of

correct ones.

The difference between the old and new systems is subtle but highly significant, according to researchers who have studied the

psychology of witness identification. At present, eyewitnesses

browse through photographs of suspects, comparing, contrasting and

re-studying them at will.

Under the new system, victims and other eyewitnesses would be shown pictures one after the other. They would not be allowed to

browse. If they wanted a second look, they would have to view all

the photos a second time, in a new sequence. Also, the pictures

would usually be shown by a person who would not know who the real

suspect was.

"It's just a reality that eyewitness identifications are made under situations of incredible duress, when people are trying to

recall what someone looked like, and they can be more or less

accurate," Mr. Farmer said. "So what we're trying to do with these

guidelines is to give law enforcement a way in which we think we

can at least narrow the risk that a mistake will be made."

The new rules also change the way physical lineups, called showups, will be done, although the use of suspects and stand-ins

is so rare in New Jersey these days that some prosecutors cannot

remember the last time they were used. As in photo lineups, the new

rules require that in showups, individuals must be presented to the

witness one at a time, usually through a one-way mirror.

The New Jersey program, which is already being used in Camden and Hunterdon Counties, grows out of a quarter-century of psychological

research and is supported by recommendations published two years

ago by the United States Department of Justice for police forces

nationally.

The federal recommendations followed a 1998 study by the National Institute of Justice, a research arm of the Justice Department,

which asked police officials, defense lawyers, prosecutors and

researchers to review 28 criminal convictions that had been

overturned by DNA evidence. The study found that in most of the

cases, the strongest evidence had been eyewitness identification.

The Justice Department published a guide titled "Convicted by Juries, Exonerated by Science" in 1999, summarizing its

recommendations for change, saying, among other things, that

sequential lineups were an acceptable option.

New Jersey, working with a pioneer in the field, Gary Wells, a psychologist and researcher at Iowa State University, soon began

drawing up its own guidelines.

New Jersey's program was developed by Debra L. Stone, deputy director of operations and chief of staff in the state's Division

of Criminal Justice. Ms. Stone said that the plan elicited howls of

protest when it was introduced to county prosecutors, and local

police departments and prosecutors, who feared that the new

procedures would make it harder to win convictions because fewer

suspects would be identified.

They also expressed concerns that the procedures would impose additional burdens on the short-handed police departments. "But we had a program for them where we had Professor Wells come in to tell them some of his horror stories about misidentifications, and about

the way people's memories work, and in the end they were very supportive,"

Ms. Stone said.

Chief John Miliano of the Linden, N.J., Police Department said: "Every time you see something coming along that makes your job a

little harder, you kind of cringe a little. It's going to take

extra time and personnel, but if it's going to make a case a little

more solid or if it's going to eliminate a bad identification or a

situation where an officer may try to influence an identification,

then it's beneficial."

Both Mr. Farmer and Mr. Wells said they believed that New Jersey will be the first state in the nation to use the new lineup techniques. Over the years, researchers like Mr. Wells, and Rod Lindsay, a psychology professor at Queen's University in Kingston, Ontario,

have demonstrated that sequential lineups made a huge difference.

Professor Lindsay would stage a mock crime - like a purse-snatching - in front of a group of people who had

agreed to participate in a study. He would then show the witnesses

a traditional lineup of suspects, like a group of photographs or a

number of people standing in a row, but he would not put the

"purse-snatcher" in the lineup. About 20 percent to 40 percent of

the witnesses mistakenly identified someone as the criminal.

When the same suspects were put in a sequential lineup, and the eyewitnesses were shown photographs one at a time, and only once,

the rate of false identifications dropped to less than 10 percent.

Other experiments showed that witnesses who did remember the criminal were just as likely to pick that person out of sequential

lineups as they were from traditional simultaneous lineups.

The reason that sequential lineups work is rather simple. In simultaneous lineups, Professor Lindsay said, witnesses are able to

compare individuals, choosing one from the group who looks the most

like the person they think they saw commit the crime. But a

sequential lineup limits the ability to compare.

The psychologists think that the chance of misidentification is reduced the most by allowing witnesses to view photos only once.

New Jersey, however, plans to let witnesses see photos more than

once, although the sequence would be changed between viewings. And

even if witnesses declare a decision in midsequence, they are

required to view the sequence through to the end, to assure that

each picture has been seen the same number of times.

Harold Kasselman, deputy first assistant prosecutor in Camden County, which has been using the new system since December, said,

"Our feeling is that if they request it, we shuffle all eight

photographs again and show them again in random order." A witness

who makes an identification is told to sign and date the chosen

photo, and to initial the other seven. All eight photos become

evidence in the case.

Another crucial innovation, the researchers found, was to be sure that a neutral third party conducted the lineup, in what is called

a blind test. If the detective knows which person is the suspect,

it could allow the detective, consciously or not, to guide the

witness.

"Let's say you're the detective and you've got your person in position three" in the group of photographs, Professor Wells said.

"You show this spread to the witness and the witness says, `Well,

No. 2.' A natural reaction is to say, `Be sure you look at

all the photos.' On the other hand, if the first words to come out

of the witness's mouth are, `No. 3,' then it's, `Tell me

about No 3.'

"It's just a natural human reaction," he said.

The studies also showed that witnesses can be just as

certain about a mistaken identification as a true one.  And being told that a false

identification is correct makes witnesses even more certain.

"It is one thing to detect lying in court, but how do you figure out that one person made a mistake in identifying a suspect and the

other didn't?" Professor Lindsay said. "Both are perfectly sincere

in telling you the truth as they know it."

But even though the experts are confident that they have found a better way to conduct lineups, they have had a difficult time

convincing law enforcement officials.

Attorney General Farmer said that New Jersey is unusual in that he has the power to order a change in lineup procedures statewide. In New York's less centralized law enforcement network, however, officials say that a change to sequential lineups would most likely

need to be spearheaded by district attorneys, but in cooperation

with the police and the attorney general. District attorneys said

that while they were interested in whether sequential lineups might

improve identifications, the matter needed far more study and

debate before a shift could be made.

George A. Grasso, the New York City Police Department's deputy commissioner in charge of legal affairs, said group lineups were

based on long-established case law and could be particularly hard

to change in New York's sprawling system.

New Jersey's new rules would allow an investigating officer to conduct the lineup in cases where no neutral officer is available

because the police department is so small, or because it is so late

at night.

Still, as Chief Miliano pointed out, detectives talk among themselves about their cases all the time, so even a fair-sized

department like his might have a hard time finding an officer with

no knowledge of a given case to conduct the lineup.

But as Richard P. Rodbart, deputy first assistant prosecutor for Union County, said, police officials know that once the new

guidelines have fully gone into effect, any other approach will

become a liability that defense lawyers will pounce on.

"I don't want an officer getting on a witness stand after he's used the old way and being asked, `By the way, sir, are you

familiar with the order from the attorney general that there has

been a new way to do identifications?'" Mr. Rodbart said.

"And then the officer says, `Yeah, I heard something about that.'

And then the defense attorney's voice rises, `Did you follow that

order?' and bang, he's on track to knock the case down."