Psychology is at its most interesting to me when it demolishes what we believe to be true on the basis of common sense, and it does this with alarming regularity. Take our memories. The act of remembering is something we do all the time, so we feel we have an innate understanding of how our memory works. But it is precisely this familiarity that leads us astray. Except for moments where we forget where we placed the keys, we are not privy to the many ways in which our memories let us down. Psychological experiments, however, can make those failures clear, and they have revealed that our memories are more incomplete, inaccurate and easily changed than we would like to think.

Daniel Simons from the University of Illinois and Christopher Chabris from Union College Schenectady have done a large survey to look at our misconceptions about memory. They asked a nationally representative sample of over 1,800 Americans to say how much they agree with various statements, and compared their answers to a small group of experts – professors, polled at a psychology conference, who had been studying memory for more than 10 years. This slideshow shows what they found.

On the whole, 60 percent of people agreed with statements that the experts almost totally rejected. These misconceptions can have severe consequences, when they influence the outcomes of court cases. As Simons and Chabris write, “This discrepancy between science and popular beliefs confirms the danger of relying on intuition or common sense when evaluating claims about psychology and the mind. Accordingly, scientists should more vigorously communicate established and uncontroversial results (alongside new and surprising findings) in a way that leads to broader public understanding.”

Would you send someone to jail on the basis of video footage shot with a low-resolution camera whose lens has dirty marks around the sides and a massive hole in the middle? Probably not, and yet that is basically what eyewitness testimony is. While it looks like we see the world in vivid detail, like the display on a high-definition television, that’s largely because the information from our eyes is heavily processed by the brain. It covers the missing information in our blind spot, and smoothes over the lack of detail around the edges. It’s a filtered version of reality.

This is hardware problem, but the software has glitches too. Our view of the world is sensitive to our expectations, our desires and where we assign out attention. Simply put, we see what our brain wants us to see. The camera metaphor implies a passive process where we switch on our memory, and it dutifully records away. The reality is very different.

As we’ve seen, the memories of eyewitnesses can be fickle things. But confident eyewitnesses can sway the minds of juries. There is a grain of sense in this – look at a large group of people and they’re generally more accurate if they’re more confident in their memories. But for any individual, confidence is a poor gauge of accuracy because we all differ in how confident we are in the first place. The consequences of overly relying on the confidence of memory can be catastrophic. Eyewitness misidentifications are the single greatest reason why innocent people are convicted for crimes they did not commit.

People often imagine their memories to be like vast libraries, where information is written down, filed away, and then brought back when it’s needed (or lost in some dusty shelf). But the act of remembering is more complicated than that.

Every time we bring back an old memory, we run the risk of changing it. It’s more like opening a document on a computer – the old information enters a surprisingly vulnerable state when it can be edited, overwritten, or even deleted. It takes a while for the memory to become strengthened anew, through a process called reconsolidation. Memories aren’t just written once, but every time we remember them. This might be useful in terms of conquering traumas and phobias, but it’s much less helpful in a courtroom.

People generally believe that even if we focus our attention on a task, we will be distracted by surprising things going on around us. Time and again, Simons and Chabris have shown that this isn’t true. If you are one of the few people left who hasn’t heard of their famous illusion, try this or one of many other similar tests. On the surface, this seems to be a misconception about attention rather than memory, but obviously, we only explicitly remember things that we see in the first place.

In courtrooms, a person’s honesty can be called into question is it’s deemed that they should have noticed something obvious going on around them. Such a case happened in 1995. While chasing after a suspect, a Boston policeman called Kenneth Conley ran directly past several other officers mistakenly beating another man. Conley said he didn’t see the beating, to great disbelief. He was indicted for perjury and obstruction of justice, and sentenced to 34 months in prison. Through a staged experiment, Simons and Chabris showed that it’s entirely possible to run past a vivid beating if your mind is on other things

This one has few practical implications, but it is interesting nonetheless. Amnesia is often used as a convenient plot device in films and TV, where people suddenly lose all memory of their names or the past lives. This can happen – it’s called a fugue state, but it’s very rare. The film Memento has a more accurate portrayal of amnesia, or at least the variant known as “anterograde amnesia”. As in the film, people with this condition lose the ability to entrench new memories, losing new information and experiences after a short space of time. But even in this film, the hero loses a sense of his own identity, something that happens very rarely in real life.

People can also suffer from “retrograde amnesia”, where they lose older memories, particularly those that happened immediately before an accident or injury. It typically occurs alongside anterograde amnesia, and is rarer as a stand-alone condition.