New Interview Tools Help Crime Victims, Witnesses Recall More

At first, the woman couldn’t remember if the suspicious-looking man she saw near the murder scene wore any jewelry.

But a few minutes later, as she concentrated on her recollection of him turning his head to the side, she remembered a gold earring.

That memory helped solve the case, said psychologist Ronald Fisher, who interviewed her at the request of police. The payoff came from an experimental technique called the cognitive interview, aimed at getting crime victims and witnesses to remember more than under standard police questioning.

The technique is one of several contributions psychology researchers hope will help solve crimes and convict criminals.

Much of this work has been done only in laboratories. But the cognitive interview recently showed promise in a real-world test with detectives at the Metro-Dade Police Department in Miami.

It “proved to be real fruitful,” said Police Sgt. Jim Wander. “The guys who took part in it are appreciative of the results, and I think they’re probably better detectives overall as a result of it.”

Results were reported in the October issue of the Journal of Applied Psychology by Fisher and Michael Amador of Florida International University and R. Edward Geiselman of UCLA.

By analyzing interviews with robbery victims and witnesses, they found that seven detectives trained in the technique elicited 47% more information than before training, and 63% more than did nine untrained detectives.

The technique involves asking fewer questions than in a standard interview, putting more burden on the witness to actively generate information rather than simply respond to queries, Fisher said.

Among its strategies is discerning the witness’ job or hobby interests, which might indicate details he is more likely to recall. A clothing designer might be able to describe in detail what a suspect was wearing, for example.

The witness is also asked to recall his emotions and thoughts at the time of the crime, because studies suggest that will help recollection. That can be tricky in the case of a victim who was frightened by the crime, Fisher said.

The interviewer also asks the witness what happened, not so much to get information as to determine the witness’ vantage point in seeing the crime and the “mental pictures” he has of it. The meat of the interview lies in asking the witness to recall the event, one mental picture at a time, and probe each picture for detail as it unfolds.

For example, if a suspect initially approached the witness and then walked away, the witness should be asked about the suspect’s face only while recalling the approach. Questions about the back of the suspect’s jacket should wait until the witness has moved on to recalling the departure.

The technique has “a lot of promise,” said Gary Wells, chairman of the psychology department at Iowa State University and the author of a handbook on eyewitness identification for police.

But he and Elizabeth Loftus of the University of Washington have some reservations. For example, the interview combines several techniques, and it is not clear which produces the results. Finding that out might help researchers determine when to use the interview and when not to, Wells said.

Loftus also questioned whether the extra details recalled during the cognitive interview are important. “Some of them are important, some of them are not,” Fisher said. “I go for all the information I can get.”

Wells said more tests of the technique are needed.

Another line of memory research deals with the risk that a victim or witness will falsely identify an innocent person. Several hundred people a year may be falsely convicted of felonies in the United States through erroneous identification, said Ken Deffenbacher, a psychology professor at the University of Nebraska at Omaha.

Laboratory experiments have shown two strategies useful in reducing the risk of false identifications in police lineups, Wells said. Lineups include a person suspected by police and several people who generally resemble him.

In one strategy Wells studied, the witness is shown a “blank” lineup which, unknown to him, does not contain any suspects. If he fingers somebody, he is discarded as unreliable.

In the other strategy, the witness is shown one person at a time and asked if it is the right person. In the laboratory, this technique reduced the incidence of false identifications by 20% to nearly 40%, Wells said.

But Lt. Ron Lewis of the robbery-homicide division of the Los Angeles Police Department said he considers the “blank” lineup unnecessary because a regular lineup can already expose an erroneous witness if the witness fails to choose the suspect. He also said the district attorney generally will not file a complaint against a person identified by only one witness; corroborating evidence or multiple witness identifications are generally needed.

On another aspect of eyewitness reliability, studies show that people do more poorly in identifying people of a different race than of their own, said psychologist John Brigham of Florida State University. His recent analysis of 14 studies involving blacks and whites found the problem affected each race about equally. “A lot of people had assumed that whites would do worse with blacks than blacks would do with whites,” Brigham said.

The reason for the problem is not clear. The observer’s degree of racial prejudice does not seem to make a consistent difference, and evidence suggesting that the cause is lack of exposure to people of a different race is “pretty weak,” he said. Perhaps scientists need a better measure of exposure, including types of exposure, to demonstrate such a relationship, he said.

Brigham suggested that in appropriate cases, jurors be told of the hazards of cross-racial identification. He also recommended that lineups be arranged by an officer who is the same race as the suspect and may be better able to choose people who resemble the suspect, providing a fairer test of the identification. Police departments do not appear to do that generally now, he said.

Gerald Arenberg, executive director of the National Assn. of Chiefs of Police, said he doubts that such a practice would be widely used because of the logistics of having enough officers of the proper race available.

Wells said that when he describes his lineup techniques and the cognitive interview while giving workshops to police, “they are quite impressed with the kinds of findings that we’re coming up with.

“I really think this is one of the areas of applied psychology that is making a real contribution and a real difference.”