It was May 21, 1991, and it was a rainy night in New York City. I found myself sitting in a rather cramped and dreary backroom of “The Joan Rivers Show” with my friend, the brilliant Dr. Elizabeth Loftus.

Dr. Loftus had been invited to appear on the late night talk show to promote her groundbreaking book, “Witness for the Defense: The Accused, The Eyewitness, and the Expert Who Puts Memory on Trial.” “The Joan Rivers Show” had flown her in from Seattle, where she was a renowned psychology professor at the University of Washington. As a researcher, she had made astounding breakthroughs in demonstrating what most of us in the criminal justice profession already knew—that eyewitness testimony was not as reliable as people believed. “The Joan Rivers Show” had suggested she come with a lawyer with whom she had worked; to my profound delight, she had asked me to accompany her.

That particular night, one of the late night comedian’s guests also happened to be one of Rivers’ closest friends, the actress Mariette Hartley. As their conversation warmed, the host decided to keep Hartley on for the final segment and bumped us completely off the show. Elizabeth was not pleased. She had interrupted her busy schedule and flown across the country for nothing.

Sign Up for E-News

But, before Dr. Loftus flew back home, she did make the time to address a large assemblage of lawyers from my office. The lecture part of her presentation put forward the results of her tedious but important research. According to her findings, recalling an event is a tri-part event. First, when an event occurs, the quality of our perception can be greatly influenced by myriad factors, many of which can dilute our ability to accurately record what is happening. Second, there is a process through which we imprint what we have just experienced in our memory banks. That process can also be highly suspect. Third, when called upon to retrieve this information, the method employed can have a dramatic effect on what we actually recall. To illustrate her point, she showed the group of over 100 defense attorneys a series of slides and then asked us to fill out a questionnaire. Through the use of suggestive questioning, the professor was able to convince over 70 percent of the lawyers that there was a red barn in one of the slides. Of course, there was no barn, red or otherwise, in any part of her presentation. Her point was firmly established.

After her disappointing non-appearance on “The Joan Rivers Show,” Loftus flew back to  Seattle and continued her research. Since 1991, she has written several other books. Her work in the field of identification testimony has sparked a movement, demanding long overdue reforms in our criminal justice system and has given birth to a new generation of psychologists who have picked up her torch.

Recent strides in utilizing D.N.A. testing in closed cases have led to 364 overturned convictions nationwide. The vast majority of them have been labeled misidentifications. One case that the New York  Court of Appeals sent back for retrial was the case of Otis Boone. Mr. Boone had been convicted in 2011 of two robberies based solely on eyewitness testimony. He is black and the victims are white. The defense in that case wanted the jury to be informed of what psychologists’ research has been saying for years—that witnesses often struggle to identify strangers of a different race.

The trial court would not allow that information to be conveyed to the jury, and Mr. Boone was convicted. The Court of Appeals found the trial judge’s decision alarming and reminded all of us that, “Mistaken eyewitness identifications are the single greatest cause of wrongful convictions in this country.” At his second trial last month, thanks to the skillful work of my brilliant and dedicated colleague, Bess Stiffelman, Mr. Boone was acquitted. Through skillful defense work, the jury was made aware that Mr. Boone had used his public benefit card a mile away from one of the robberies only five minutes before it occurred. That fact, taken together with the doubts about identification, must have been persuasive because it only took the new jury five minutes to find Mr. Boone not guilty.

One of the key witnesses in the second trial was another brilliant psychology Professor, Dr. Nancy Franklin, who teaches at Stony Brook University. I have consulted with her on many cases and I’ve always been extremely impressed. She was gracious enough to speak to me as I prepared this column:

Can you tell us about your concerns surrounding eyewitness testimony?

When we listen to witnesses in a courtroom we expect, especially when it comes to important matters like identification, 100 percent reliability. But we may be asking for too much. We are requiring someone to expound on an event that may have occurred two or more years ago under very stressful situations.

Can you describe some of the factors that may lead to a false identification?

Weapon focus, multiple perpetrators, cross racial identity, partial disguise, the speed of the events; all can contribute to an inaccurate identification.

The expert witness for the prosecution in the Boone case, Dr. John Wixted, suggested that the witnesses’ “confidence” in making the identification should have been the key factor, not cross racial identification issues. Is he correct?

Confidence in one’s identification is not a filter against misidentifications for many reasons. For example, in the Boone case, the composition of the lineup was highly unfair. There were terrible fillers (people placed in lineups other than the suspect) who looked nothing like the defendant. When you couple that with the biased instructions given by the police to the victim, you are left with a highly suggestive affair. To make matters much worse, the officer told the eyewitness that “he needed him to be sure.” The result was that the witness became locked into and extremely confident in what was a complete misidentification. This was a perfect example of a situation where feedback from a law enforcement officer can lead to an artificially manufactured confidence which only becomes more rigid every time the victim is asked to go back over the events.

No system is perfect. Eyewitness testimony continues to be a major source of information in our criminal justice system. However, thanks to the continued work of people like Elizabeth Loftus and Nancy Franklin, we are much more cognizant today of the need to carefully scrutinize the myriad of issues surrounding both the experience of the victim during the incident and the subsequent identification procedures themselves. Mr. Boone spent approximately seven years in jail for crimes he did not commit. If nothing else, his case demonstrates the need for nothing less than a full commitment from all of us in the criminal justice system to avoid wrongful convictions at all costs.