Human memory has been shown again and again to be far from perfect. We overlook big things, forget details, conflate events. One famous experiment even demonstrated that many people asked to watch a video of people playing basketball failed to notice a person wearing a gorilla suit walk right through the middle of the scene.

So why does eyewitness testimony continue to hold water in courtrooms? A new nationwide survey of 1,500 U.S. adults shows that many people continue to have the wrong idea about how we remember—and what we forget.

Here are four common incorrect assumptions about memory, held by some of the survey subjects, that experts say should be forgotten:

1. Memory works like a video camera, recording the world around us onto a mental tape that we can later replay.

Nearly two thirds (63 percent) of those in the random telephone survey said that they agreed with this model of a passively recorded memory. This notion runs counter to research that has shown events to be recalled based on "goals and expectations," the researchers behind the survey wrote in a new paper, published online August 3 in PLoS ONE. It also "contradicts the well-established idea that memory retrieval is a constructive process," too, which can be shaped by assumptions and beliefs, noted Daniel Simons, of University of Illinois, and his co-author, Christopher Chabris, of Union College, both of whom are psychology professors.

2. An unexpected occurrence is likely to be noticed—even when people's attention is elsewhere.

More than three quarters (77.5 percent) of people thought that this would be the case. Clearly, they are unfamiliar with the gorilla suit study. That work and other research have shown that unexpected—and even preposterous—details frequently go unnoticed, and thus do not make it into memory. Aside from a false certainty that one would notice more people wearing oversized primate costumes, this presumption could have some serious implications for the legal system and eyewitness testimony. "If juries and lawyers believe that a suspect 'should have' noticed some event, they will tend to see claims of ignorance as deliberate attempts to deceive," Simons and Chabris wrote.

3. Hypnosis can improve memory—especially when assisting a witness in recalling details associated with a crime.

Most memory experts disagree with this statement, but more than half (55.4 percent) of the surveyed public thought that it was accurate. Courts have already steered away from accepting testimony that was gathered through hypnosis. And many studies have demonstrated that people under hypnosis—and even those who are not—can often be led by questioners to "recall" things that never occurred.

4. Amnesia sufferers usually cannot remember their identity or name.

Although soap operas might lead you to conclude otherwise, most common forms of amnesia interfere with the formation of new long-term memories—usually as a result of a major brain injury. The researchers cite the movie Memento as a reasonably accurate portrayal of the condition, but most popular portraits "depict amnesia as something more like a much rarer fugue state in which someone cannot remember who they are and suddenly take leave of their home and work," they noted. Perhaps because of the prevalence of this blank-stare amnesia in television and movies, a whopping 82.7 percent of those surveyed shared this (incorrect) view of the condition.

The survey also found that nearly half (47.6 percent) of respondents said that once a memory is formed, it is set in stone. This is also not true, say the researchers: "Our memories can change even if we don't realize they have changed," Simons said.

Along these lines, more than a third (37.1 percent) of people thought that "confident" testimony from a witness should be adequate for a criminal conviction. However, many defendants who were later shown to be innocent via DNA testing had originally been convicted based on a faulty ID by an eyewitness. And as the researchers pointed out in their paper, being confident about your memory of an event is a good predictor of its actual accuracy, but "the link between confidence and accuracy across individuals is more tenuous, in part because people differ in their baseline levels of expressed confidence."

A lesson to be gleaned from all of this might exonerate a group that might need all of the credibility it can get these days: politicians. "The extent of these misbeliefs helps explain why so many people assume that politicians who may simply be remembering things wrong must be deliberately lying," Chabris said. But imperfect memories alone, of course, do not guarantee anyone is always striving to be deception-free.

But if there's one thing to remember about the findings, it's that "people tend to place greater faith in the accuracy, completeness and vividness of their memories than they probably should," Simons said.

To see how you and other readers measure up to the experts, the authors of the study (who also wrote the book The Invisible Gorilla) have created an online quiz that also shows the rates of correct—and incorrect—responses to the survey questions.

Image courtesy of iStockphoto/DebbiSmirnoff