Study Shows How Video Can Alter Eyewitness Memory

Wednesday, September 23rd, 2009

Wired reports on a fascinating new study showing how eyewitness memory can be influenced by video. Participants in the study were paired with a partner (who was actually part of the research team) to play a gambling-based computer game. The participants bet money on their own ability to answer multiple choice questions. The game relied on participants to honestly pay back money they’d earned when they got a question wrong.

After they had finished with the questions, the participants were told that their partner had cheated, even though the partner hadn’t. One group of students were then shown a video that had been digitally altered to make the partner look as if he had actually cheated. Though they were told to only report their partner if they were 100 percent sure he had cheated, and that the partner would be punished based on their decision, about half the participants who viewed the video still reported their partner for cheating. The second group wassn’t shown the video, but only told of its existence. Just 10 percent of them still reported their partner for cheating.

The article may overstate the study’s lessons a bit, at least in their relevance to the criminal justice world. It seems unlikely that many criminal cases are tainted by video digitally manipulated to show obvious guilt—even if you’re cynical enough to believe law enforcement officials would try it, I’d think it would be pretty easy for someone with expertise in video editing to detect. But the broader point—that eyewitness memory is highly susceptible to suggestion—is worth heeding, and has been confirmed in numerous other studies.

Digg it |  reddit | |  Fark

15 Responses to “Study Shows How Video Can Alter Eyewitness Memory”

  1. #1 |  C. S. P. Schofield | 

    I’ve read of dozens of this kind of study, and each one has always seemed to me to be grounds for pounding the snot out of whoever thought them up.

    “We’re going to lie to you, but it’s OK, because it’s for Science!”


    Is it just me?

  2. #2 |  ClubMedSux | 

    During the summer I spent at Northwestern University’s Center on Wrongful Convictions (yes, a shameless plug; send them a donation anyway!) I learned a lot about the unreliability of eyewitness testimony. That being said, I think this study actually cuts both ways. Yes, if you have a situation where somebody fabricates evidence, it can manipulate one’s memory. However, my experience (limited as it may be) is that the more likely scenarios in a crime investigation are: 1.) the witness has a distinct memory that is contradicted by physical evidence–and the physical evidence is NOT fabricated and IS more reliable than the witness’s memory, or 2.) the police present a witness with false information in an attempt to elicit information (often a confession). What this study tells me is that, in the first scenario, people are more likely to trust the physical evidence (in this case a video) over their own memory, and, in the second scenario, 90% of people will dismiss second-hand evidence they can’t verify. Both seem like reasonable responses to me, and preferable to the alternatives.

  3. #3 |  freedomfan | 

    I agree that this is just more of a sign that people don’t trust their own memories when confronted with what they believe to be reliable physical evidence.

    The scary thing is that police routinely lie to witnesses about the evidence they have collected in order to get testimony from the witness. I think studies like this back up the idea that police should not be allowed to do that because it can alter a witness’ memory of what really happened. In other words, if 10% of the people reported certainty that their partner cheated when simply told there was video evidence of it (even if those people hadn’t seen the video), then there is an obvious mechanism for police to generate false testimony by telling a witness they have hard evidence (video or otherwise) of something that they don’t. Basically, police lying to a witness during an interrogation is a method of fabricating false evidence and the practice should be banned and evidence gathered that way ruled inadmissible. The 10% false certainty result of this research might have been far higher if it had been conducted using the high pressure techniques police do in their interrogations.

    I also think this sort of thing (holding press conference where they imply they have hard evidence of guilt, etc.) is a method that prosecutors routinely use to garner public support for a prosecution. Heck, don’t prosecutors routinely open their cases in front of juries by saying that they have evidence that will show conclusively that the accused is guilty?

  4. #4 |  ChrisD | 

    C. S. P. Schofield,

    Maybe it’s not just you, but I don’t think that this is too bad. It’s not like they told them personally traumatizing information or anything. This study has given us a valuable insight into human memory that is hard to imagine us discovering without any deception. I’m biased, since I’m a cognitive neuroscience researcher, but I never have reason to use deception in my studies so it’s not.

    It’s actually astonishing just how malleable memory is. There’s an experiment where volunteers are asked to view video of a car accident, then fill out a questionnaire about it. Different people are given different questionnaires. People whose surveys asked “How fast were the cars traveling when they ‘crashed?'” estimated higher speeds than those who were asked “How fast were the cars going when they bumped?” Same memory going in, different one coming back out depending on how it’s elicited.

    There are loads of studies like this. I think I’ve taught almost a full class worth of material on it.

  5. #5 |  ChrisD | 

    ….but I never have reason to use deception in my studies so it’s not like it hits too close to home.


  6. #6 |  Bill Anderson | 

    As an economist, I hardly am surprised to see government agencies (police and prosecutors) use extremely flawed means to “solve” crimes and get convictions. For that matter, prosecutors in Mississippi still are using Steven Hayne as an “expert” witness even though the man is a fraud.

    Look at the fraud that prosecutors and police perpetrate all the time in investigating crimes. Should we be surprised when they resist using real science instead of junk? They are government employees, after all.

  7. #7 |  Jim Collins | 

    I used to work in aviation. For a while I wanted to be an air crash investigator. I was actually taking classes and training towards that goal. I was invited to a seminar given by the NTSB concerning the care and handling of witnesses. One of the questions asked was “You have a choice of two witnesses. One has a pilot’s license and twenty years in aviation. The other is a 21 year-old waitress from Denny’s. Which has the potential to be the better witness?” Everybody picked the pilot, which was the wrong answer. It was explained to us that the pilot’s aviation background taints his testamony, because he knows too much about aviation. He may inadvertantly overlook critical information because he thinks that it wouldn’t be useful and he would try to describe what HE thinks happened. The waitress, on the other hand, would give a more accurate description, because she doesn’t know anything about what may have caused the incident.

  8. #8 |  John Wilburn | 

    These types of studies seem to confirm the adage, “Figures don’t lie… But liars can figure.”

  9. #9 |  omar | 

    @#1 | C. S. P. Schofield

    In studies like this, you have to do a couple things to use deception like…

    1) get it approved ahead of time by a board of folks who determine if the test could harm the test subjects in any way
    2) state explicitly “deception may be used”

    Often times, like in this study, the “study” the participants believe is happening is a placebo for the participants while the real experiment is set up around the false study.

    This is most often done to get an honest reaction out of the study participants. When the participants understand the subject is “cheating”, they may cheat all the time or never at all to skew results or make themselves look good.

  10. #10 |  z | 

    So far we’ve learned that a. expert testimony is bogus b. forensic science is bogus, c. confessions are usually bogus, and now d. eyewitness testimony is bogus. Audio and video can be manipulated. The agitator is rapidly coming to the position that there is no valid way to prove a case beyond reasonable doubt.

  11. #11 |  Radley Balko | 

    So far we’ve learned that a. expert testimony is bogus b. forensic science is bogus, c. confessions are usually bogus, and now d. eyewitness testimony is bogus. Audio and video can be manipulated. The agitator is rapidly coming to the position that there is no valid way to prove a case beyond reasonable doubt.

    Not quite. What you’ve learned is that forensic science can overstate its conclusions, and that some parts of it (bite mark testimony, for example) are bogus). You’ve learned that confessions don’t always indicate guilt, not that they’re “usually bogus.”

    The point isn’t that there’s no way to prove a case, it’s that we shouldn’t let the state overstate its case. Prosecutors still win 80-90 percent of the cases that come to trial. The Innocence Project estimates that 4-6 percent of those convicted of felonies are innocent. Maybe with more skepticism and better rules about what can be admitted, and the degree of certainty with which experts can testify, the innocence numbers will drop.

    The point is to be more skeptical about evidence, not to dismiss it entirely.

  12. #12 |  Billy Beck | 

    “The point is to be more skeptical about evidence, not to dismiss it entirely.”

    It’s a caveat, Radley, that turns up quite a bit too rarely in these sorts of articles. “The point” often arrives a bit more pointed than I think is useful, especially when I consider prevailing epistemologies in times like these.

    It would be a lot better to simply document the facts of police and prosecutor misconduct.

  13. #13 |  albatross | 

    Apparently, a lot of our forensics are based on best guesses and “looks right to me” sort of reasoning, rather than anything rigorous. At the same time, the way we deal with taking statements from witnesses and weighing witness testimony is mostly based on old intuitions about memory that have turned out to be wrong.

    Those are important things to notice. We need to try to move both forensics and handling of witnesses onto a better footing, because we would like our courts to come to correct answers more often. This doesn’t mean we can’t trust any evidence, it means that just because a witness truly believes he saw X doesn’t mean he actually saw X, and his memories are more suspect, the longer it’s been since he saw X and the more information he’s been given since then about the case, true or false.

  14. #14 |  kyle | 

    A sort of not-exactly related version of this is the recent beason, IL murder. The first time i heard about it, 5 people were killed from blows to the head. But now, they weren’t killed by gunshots and being stabbed. The autopsy of the brutal attact might come back showing they died from consuming drain cleaner with their breakfast cereal before not being shot and stabbed. The recent reports overwrote the initial reports, and the public will be re-horrified when the autopsy suggests blunt force trauma to the head and maybe torso. Logic in the hands of ethos dealing with eros. I wasn’t a witness to the crime, but all i need is a picture of the face of the guy driving the gray primer pickup truck and i could create a rhetorical sequence of events that proves beyond a doubt that 9/11, an event i watched all but the first 2 minutes of live on television, was an inside job. In a debate like this, the entity that can talk loudest and longest walks away the victor, regardless of the facts or the law.

    Have you ever watched a single movie a number of times, only to catch a line you’d never heard before that alters your overall interpretation of the entire movie? And i don’t mean like how george lucas reedited the first 3 star wars movies to make them fit better with the second 3. But it would give me an apoplectic stroke to be able to state the truth concerning if Han did or did not in fact shoot first. It would take some serious effort on the part of an actual fan to explain the implications of either choice, and in the end that is just one criminal case (but with better backstories and character development than the real world). Imagine the nightmare of trying to get 12 random people pulled out of their own lives to be able to reach the conclusion that both descriptions of the actions of a drug smuggler could and did very well happen. I just want to go home today, so i’ll just go with an easy “yes” or “no” instead of hanging the jury and let it become somebody else’s problem.

  15. #15 |  Mo | 

    To me the 10% that rat out their partner just because they were told that they cheater is the most disconcerting part.

  16. #16 |  Sexual Politics, the Law and the Roman Polanski Arrest | Xenia Institute | 

    […] Study shows how video can alter eyewitness memory (Read more). […]