Barry Scheck
Attorney and Founder of the Innocence Project
06:51

The Dangerous Unreliability of Eyewitnesses

To embed this video, copy this code:

Eyewitness testimony is the single greatest cause of wrongful convictions, says Scheck. Psychologists and researchers have demonstrated a variety of ways that witnesses regularly misidentify suspects.

Barry Scheck

Barry Scheck is the co-founder and co-director of the Innocence Project, a national organization that uses DNA testing to exonerate wrongfully convicted people and implements policy reforms to prevent future injustice. Founded in 1988 under the auspices of the Benjamin N. Cardozo School of Law at Yeshiva University, the Innocence Project has exonerated hundreds through post-conviction DNA testing. Scheck is also famous for having defended notable clients like O.J. Simpson, Hedda Nussbaum, Louise Woodward, and Abner Louima. 

Scheck is currently a professor of law at Cardozo and a commissioner on New York's Forensic Science Review Board, a body that regulates all of the state's crime and forensic DNA laboratories. He is first vice president of the National Association of Criminal Defense Lawyers and serves on the board of the National Institute of Justice's Commission on the Future of DNA Evidence. In 2001, along with Innocence Project co-founder Peter Neufeld, Scheck co-authored the book "Actual Innocence: Five Days to Execution and Other Dispatches from the Wrongly Convicted."

Transcript

Question: How reliable is eyewitness testimony?

Barry Scheck:  Eyewitness testimony of course is the single greatest cause of the conviction of the innocent and what is fascinating about that is that we have now 30 years, more than 30 years of fantastic research from experimental psychologists who have you know recreated crime scene… recreated events and played around with which different way we do a photo array or a lineup and different techniques that are used correlate more with accurate and reliable identifications in which techniques lead to errors and a number of reforms have been adopted by federal commissions and the American Bar Association, the International Association of Police, lots of policing entities based on really good scientific research.  I mean a lot of this research is meta analysis.  That is where you pull together lots of different studies and you have a lot of confidence that certain things are real affects and I’ll give you some examples that may surprise people. 

For example, before you show somebody a photo array or a live lineup you should tell the witness, “We’re now going to show you some pictures or people.”  “Please keep in mind that the real perpetrator may or may not be in this photo array or lineup and if you don’t make an identification don’t worry, the investigation will continue.”  Just giving that warning dramatically reduces incorrect identifications without really reducing correct identifications and that is of extraordinary importance obviously in terms of being able to go out, find the real perpetrator and not arrest the wrong person.  There is lots of reasons we think that this warning will reduce error.  Certainly it inhibits people from guessing and there kind of a natural inclination to do that.  Another reform that is of really critical importance is that the person that administers the lineup or the photo array should be double blinded.  That is to say that the person should not know who the suspect is and that is important for many reasons, not the least of which is the possibility of feedback, confirming feedback. 

What again meta analysis have shown is that if you say to somebody, “You picked out number two, good.”  Or give some kind of confirming feedback whether you’re a law enforcement person or somebody else frankly what this does is it can falsely inflate the person’s certainty that they made the correct identification, the witnesses’ belief that they had a good opportunity to observe, the witnesses’ belief that they were paying attention when the crime was committed.  All of these factors by the way, certainty, opportunity to observe, paying attention are all factors that the courts used to be using or use frankly in trying to weigh the reliability of identification when we know that let’s say some suggestive procedure has been used, so confirming feedback or even just the use of an unduly suggestive procedure can falsely inflate these so called reliability factors, certain, opportunity to observe, attention paid and the legal tests are out of whack with what the scientific evidence shows.

Question: How accurate are forensic tests that do not use DNA?

Barry Scheck: More than a year ago now the National Academy of Science came out with a report about forensic science generally, an extremely important report, a landmark report that anybody really interested in this field should look at and this report basically said look, DNA testing is the only validated forensic discipline we have and in particular when you look at pattern evidence and by this they were talking about fingerprints and tire tracks and looking at striations on bullets and trying to determine whether they came from a gun, particular gun or not, all these pattern disciplines had not be adequately validated because you would have experts coming into court and they’d say, “I’m looking at the striations of this bullet that we recovered from the crime scene and now I found this defendant’s gun and I fired a bullet from that and I’m looking at the two of them under a microscope and I see all these lines and I think there is quote, unquote, sufficient agreement.”  And by that and if I’m really a big quality assurance lab I’ll bring my buddy in and he’ll look at the microscope and he’ll say, “I think there is also agreement here and from this I can now tell you that this bullet came from that gun to the exclusion of all other guns in the universe.”  Well think about it.  Scientifically on what basis?  Do you have a database?  Can you tell us anything about the frequency of finding similar patterns if indeed you have similar patterns?  Do you have measurement area here?  Can you give us a number?  You know with DNA testing we can give you a number.  We can give you the frequency of a particular DNA profile in various different populations.  Can you give us a statistic?  And the answer if no, they weren’t giving statistics and as a matter of fact they were just saying it’s unique.  This bullet comes from that gun to the exclusion of all guns in the universe.  This is nonsense and the National Academy of Science said so and said you guys have really got to go back to the drawing board and do some basic research to find out what the answers are to that question.  Even fingerprints, which we know if you take all 10 prints and you scan it in and you put it into the database of known fingerprints you can do a pretty good job of identifying somebody, but that is not the question when you’re talking about a small latent print at a crime scene.  You get a partial thumbprint let’s say and every time you lay down a print it’s always a little bit uneven and then you compare it to somebody else’s thumbprint.  Can you say that this small latent uniquely comes from that individual?  Well that is the way the testimony was going in for years and it was without scientific basis to say that it uniquely was this person’s print.  You can’t say that.


×