The ability of a person who witnesses a crime to later pick the perpetrator out of a lineup is atrocious—right? The answer seems like a resounding “yes” if you consider some well-known and rather disconcerting information. Eyewitness misidentifications are known to have played a role in 70 percent of the 349 wrongful convictions that have been overturned based on DNA evidence (so far). Psychologists have learned a lot about why such errors happen. With surprising ease, for example, participants in a memory experiment can be led to believe that they saw a stop sign when they actually saw a yield sign or that they became lost in a shopping mall as a child when no such experience actually occurred. In much the same way, an eyewitness can be led to falsely remember someone committing a crime that was actually committed by someone else.
But there is more to the story. Consider the important, and often overlooked, distinction between malleability and reliability. Just because memory is malleable—for example, it can be contaminated with the trace of an innocent person—does not mean that it has to be unreliable. What it means is that the malleability of memory can harm reliability. Once this fact is appreciated, then proper testing protocols can be put in place to minimize the likelihood that the original memory trace is contaminated.
Current procedures for collecting and assessing evidence from eyewitnesses are often not designed to minimize contamination.This problem does not apply to other kinds of forensic evidence. Imagine if police let unauthorized people have willy-nilly access to a crime scene that is under investigation. What would that mean for the reliability of the blood or fingerprint evidence, for example, collected at that scene? Blood and fingerprint evidence, per se, would not be deemed unreliable. Instead evidence collected at the contaminated crime scene would probably be declared inadmissible. Not so with eyewitness memory.
On supporting science journalism
If you're enjoying this article, consider supporting our award-winning journalism by subscribing. By purchasing a subscription you are helping to ensure the future of impactful stories about the discoveries and ideas shaping our world today.
Any evidence can potentially be contaminated, including what is considered to be the gold standard of forensic evidence: DNA. Like eyewitness memory, DNA evidence can be contaminated with the trace of an innocent person. If the contaminated evidence is relied on to establish guilt versus innocence at a trial, the risk of a wrongful conviction is high. Consider, for example, the case of Gary Leiterman, who, in 2005, was convicted of murder and sentenced to life in prison following a cold case investigation in which his DNA was found on the clothing of a woman named Jane Mixer, who was murdered in 1969. Many believe that this conviction was based on contaminated DNA evidence.
What facts gave rise to the belief that Leiterman may have been wrongfully convicted based on contaminated evidence? First, another DNA profile—one belonging to John Ruelas, who was just a four-year-old preschooler at the time of the murder—was found on a blood spot taken from the victim’s left hand. Even though no connection between Leiterman and Ruelas was ever established, the prosecution theorized that all three must have been together after midnight at the murder scene in 1969, with the preschooler bleeding on the victim for some unknown reason while Leiterman killed her. According to the same theory, 33 years later, in 2002, in an almost inconceivable coincidence, evidence samples from Mixer, Ruelas and Leiterman just happened to be together again in the Michigan State Police Laboratory. The murder victim’s cold case evidence was there because the case had recently been reopened; the 1969 preschooler’s DNA sample was there as part of an active murder investigation, and Leiterman’s DNA sample was there because he had recently been arrested for forging a prescription. The nearly simultaneous analyses of evidence from these three cases in the same crime lab in 2002—a reunion, of sorts, among Mixer, Ruelas and Leiterman, who were ostensibly last together on the night of the murder in 1969—was either an incredible coincidence or the Mixer evidence was contaminated with DNA from both Leiterman and Ruelas.
Instead of concluding that DNA evidence is inherently unreliable because of the contamination that apparently occurred in this case, a more reasonable conclusion would be that for DNA testing to be reliable, proper protocols must be followed. Few would doubt that under such conditions, DNA evidence is highly reliable. The same is true of eyewitness memory: memory can be contaminated with the trace of an innocent person, but under proper testing conditions, eyewitness evidence is highly reliable. As with DNA evidence, eyewitness evidence needs to be safeguarded against contamination.
To do this, proper testing protocols that reduce chances of contamination need to be followed. Some elements include the following: First, and most important of all, because the test itself contaminates memory, only the initial memory test provides uncontaminated results. Subsequent memory tests, including the dramatic one that occurs in court in front of the jury, constitute contaminated evidence. Second, the police lineup has to be fair (that is, the suspect should not stand out). And third, the confidence expressed by the eyewitness following an identification of someone from the lineup must be recorded. Assessing confidence is critical because it provides direct information about the trustworthiness of the uncontaminated ID. An initial eyewitness identification made with low confidence indicates that even though memory was not contaminated, the ID is untrustworthy (that is, by indicating low confidence, the eyewitness is effectively saying, “There’s a good chance that I’m making an error”). In contrast, a high-confidence ID is highly accurate, a surprising fact that has only recently come to be appreciated by experimental psychologists. In a recent review of the literature, the authors reported across 15 experiments, suspect identifications made with high confidence were, on average, 97 percent accurate!
Although the high accuracy of an initial ID made with high confidence is important to appreciate, the low accuracy of an ID made with low confidence may be even more important to appreciate. To see why, let us revisit those DNA exoneration cases that so often involve eyewitness misidentification. University of Virginia law professor Brandon Garrett analyzed trial materials for 161 DNA exonerated individuals who had been misidentified with high confidence by one or more eyewitnesses in a court of law. By itself, that fact only shows that contaminated memory is unreliable (just as contaminated DNA evidence is). But in 57 percent of those cases, it was possible to determine what happened on the initial (uncontaminated) memory test. For every one of those tests, the eyewitnesses were, at best, uncertain.
Attributing that error to the unreliability of eyewitness memory is, in our view, pointing the finger of blame in the wrong direction. Eyewitness memory is reliable when initially tested using proper procedures, but the legal system nonetheless habitually relies on unreliable (contaminated) eyewitness evidence from later IDs. The sooner police, prosecutors and judges understand that fact, the better off we all will be—including you, if you are ever incorrectly fingered by an eyewitness.
