Do you believe that memory works sort of like a video camera, faithfully recording your experiences so that you can go back later and revisit those memories, captured in pristine condition? Do you believe that if something unexpected walked into your field of vision you'd notice? Can forgotten memories be recalled through hypnosis?

If you're like 50-80% of the US population, you might answer yes to each of those questions, but you'd be wrong. That's what cognitive psychologists Dan Simons (web, blog, twitter) and Chris Chabris (web, twitter) discovered after giving a survey about memory and other psychological constructs to a nationally-representative sample of 1838 Americans.

If you've been reading my blog for a while, you may remember that the very same Simons and Chabris wrote the fantastic book The Invisible Gorilla, about the ways our intuitions deceive us. I wrote, "If there is one book that every human should read, it is The Invisible Gorilla."

In a new paper published this week in PLoS ONE (available free here), Simons and Chabris report on the results of that survey. I used this as an opportunity to speak to Dan about his research.

What made you decide to conduct this sort of study in the first place? Was it specifically for The Invisible Gorilla, or was it something you wanted to do anyway?

Our motivation really was to supplement the research we had done for our book. As you know, the book focuses on everyday illusions, cases in which people's intuitive beliefs about how the mind works are faulty. We had pretty good evidence to start with that people harbored some misconceptions about the mind, but we realized that nobody had ever measured how widely help those beliefs are, at least not on a National scale. There have been many surveys assessing the prevalence of beliefs in pseudoscience, ESP, and other hokum. But, to our knowledge, there have been relatively few surveys designed to assess how familiar people are with well established scientific findings in psychology.

Our PLoS ONE paper reports the results for the survey items about memory, but the larger survey asked about other aspects of the mind as well (e.g., the 10% myth, the Mozart effect). We may put some of those into another paper.

One of the interesting findings in the paper was that education was negatively correlated with agreement on the items in the survey - that is, as education level increased, agreement with the survey items decreased - though even those who had been in grad school overwhelmingly agreed with the statements as well. This seems to suggest that education in general isn't enough to overcome these common misconceptions regarding memory - so what do you think it will take?

That's a great question, and that finding is the one I find both reassuring and troubling. It's reassuring because it suggests that we're all in this together. To the extent that education is correlated with other individual differences like intelligence, the finding suggests that these misconceptions are not due purely to individual differences and they likely reflect something more fundamental about how we see and interact with our world. To me, that's a positive. The negative aspect is that education alone doesn't seem to eliminate these misconceptions. Of course, some aspect of education must be helpful given that the experts responded differently. What we can't tell from these results is whether the differences are due to a particular type of education (science training in general, psychology training in particular, etc.). My hope, of course, is that taking psychology courses will help reduce these misconceptions!

I think this result raises a broader issue as well. Scientists have not done a good enough job in conveying established science to the masses. Press releases always focus on the new and flashy results, not on the findings that have been established for decades. Scientists (and press officers, and the media) need to do a better job of conveying what we already know as well as what a new finding implies. If we did a better job of establishing context, then people might also be better able to examine press releases for wild new results a bit more skeptically.

You said: "Scientists have not done a good enough job in conveying established science to the masses." - I certainly agree with you on that point. How can we do better? What sorts of tools are available to scientists, researchers, professors and students in order to help bridge the connection between scientific research and public awareness?

It is a real challenge given that the media tends to glom onto the latest and greatest news rather than the established findings. I would like to see more long-form pieces about established research written in newspapers and magazines. Trade books can and do serve this purpose to some extent, and media coverage of those provides an opportunity to talk about what is well established and not just what is new. Op-Eds are another outlet that serves this purpose -- frame a current event or debate in terms of long-established research rather than the latest flashy finding. I think the biggest issue, though, is that scientists need to remember that the general public does not think about our field all the time, and most people don't have the scientific training to place new findings into the context of the established scientific literature. Moreover, few actually know the basics of psychology, even those who have taken a college course in psychology. I've attended workshops in which the presenter ran the Stroop task, and the audience was wowed. They had never seen it before. We have to remember that even the old stuff might well be new to the others, and if we can convey how well established some of that old stuff is, it might help people distinguish between the science and pseudoscience.

82.7% of your respondents had the wrong idea about the most typical form of amnesia - they thought that people who amnesia are likely to forget who they are, rather than the truth which is that they simply can't form new memories. You mention in the paper that this is perhaps due to movies like The Bourne Identity. Do you think that movies like this contribute to public misunderstanding of memory, or do you think instead that these movies simply reflect the larger misunderstandings about psychology?

I don't think we can establish cause and effect here, and both likely are to blame. Given that 80% of our respondents were unfamiliar with these principles, it stands to reason that most movie makers are as well. So, it's not surprising that they focus on the identity loss idea. Of course that also means that the idea gets reinforced in popular consciousness. So, I think they likely reflect the larger misunderstandings, but they also perpetuate them. A few filmmakers have done a good job of trying to get the details right (e.g., Memento), but most are interested in telling a story rather than conveying science.

Do you think that filmmakers have a responsibility to be more accurate about science in general, and about psychological science in particular? Should the script for The Bourne Identity, for example, have been written differently?

Although I would love to see filmmakers take a greater interest in getting the scientific details right, they aren't in the business of making science documentaries. I personally feel that there are plenty of ways to make a compelling story while also getting the facts right, but I wouldn't want to impose that constraint on artists. The book Don't Be Such A Scientist gives a great anecdote along these lines. [Neil de Grasse Tyson tells a story] about a science-related flaw in Titanic. When Dicaprio's character was hanging onto a hunk of ice, the camera showed the star-filled sky, but the sky he showed was the southern hemisphere sky. When Cameron [was confronted] and asked why he'd gotten that wrong, Cameron's quick response apparently was something along the lines of "Hmm. If I had done that, I bet the film would have grossed another $300 Million." The point is that filmmakers are far more focused on good storytelling than on making sure every detail is accurate.

Here's a video of NDGT telling this story, beginning around 1:30.

Going back to the paper, 63.0% agreed with the statement ‘‘human memory works like a video camera, accurately recording the events we see and hear so that we can review and inspect them later" - you, and I, and the experts in your paper know that this is false, but this misconception is so powerfully intuitive. It seems sort of like a visual illusion, where you KNOW how it works, but you still can't help but be fooled by it. In the case of visual illusions, this is often because the visual processing is bottom-up and not available to top-down conscious control. What do you think explains the reason for this cognitive illusion? Why is it so powerful and pervasive?

I think illusions like this one, and the other ones we discuss in The Invisible Gorilla, are pervasive because they are entirely reasonable inferences based on our daily experiences. That's also why they don't seem to vary much with education. Imagine that I showed you the gorilla video and you missed the gorilla. Now imagine that I don't tell you what you've missed. You would continue to go through life assuming that, of course you would notice a gorilla if one walked through a game. The key is that we are only aware of what we have seen, not what we have missed. That is, our experiences are based entirely on what we have noticed. And, we don't naturally think about all of the things we might have missed. After crossing the street, we don't ponder the possibility that a pink elephant might have just walked past. Doing so would waste a lot of time and effort. Because we only are aware of what we notice, we accumulate experiences of noticing, but we don't track the evidence of what we've missed. If we then reason based on our experiences, we come to the wrong conclusion about how our minds actually work.

The same principle applies for the belief that our mind works like a video camera. When we recall a personally meaningful memory, it feels vivid and detailed. We have no evidence to contradict that impression, so we trust it to be accurate. Note that the original researchers studying what they called "flashbulb memories" made the same error. They asked people to recall how they learned about the deaths of prominent political and civil rights figures, but never checked to see if those memories were accurate. Only when we are forced to check the accuracy of our memories and we happen to have documentary evidence of what really happened do we come to realize our vivid memories can be wrong. That sort of evidence typically comes only when researchers test memory in a lab setting. The exceptions to that rule come from people who do have a press pool following them and documenting their lives (e.g., political leaders). Script supervisors on movie sets also realize their memories are fallible. When I interviewed a couple of script supervisors for The Invisible Gorilla, they told me that at times, they would swear that their memory for some detail on set was accurate, but that they realized they were wrong when they checked the digital records. That experience of testing your memory and finding it lacking helps to correct the mistaken intuition that our memories must be accurate.

Were you surprised by any of the findings? Did anything make you scratch your head and wonder?

Not particularly in this case. We selected items that we suspected would lead to a big discrepancy between what people believed and the scientific consensus. And, variants of some of these items have been used in smaller scale surveys a number of times. We hoped that our large scale study would confirm the results of those smaller ones, and we thought it would. I guess one result that surprised me was that the effect of education wasn't a little bigger. We did find that people with more education tended to give answers that were more consistent with those of experts, but even the highest educated in our sample agreed with more than half the items on average. Our measure of education was fairly coarse, and we can't rule out that any effect of education was due to other factors (background, intelligence, socio-economic status, personality, etc). I was a little surprised by how many people were okay with the that a single witness should be enough for conviction. The fallibility of eyewitnesses does get a lot of coverage on legal and crime dramas. The effect of education was pretty large for that item -- nearly 50% of people without any college education endorsed that item, and the level of agreement dropped to less than 25% for those with some graduate school.

Simons, D., & Chabris, C. (2011). What People Believe about How Memory Works: A Representative Survey of the U.S. Population PLoS ONE, 6 (8) DOI: 10.1371/journal.pone.0022757

Gorilla image via Flickr/amuderick.