Myths about the brain no. 4: A person's memory of an event is very accurate

The idea that some people have better memories than others isn't a new or groundbreaking idea. But most people will feel they have a good memory with regards to events that have just happened or with events that have a personal meaning e.g. one of your birthday parties. This idea is the implicit assumption behind eye-witness testimonies; people will accurately remember what happened at a specific time and be able to give an accurate account of it.

However, there has been evidence to suggest this is not the case. People generally do not remember specific details well; they remember the critical ideas of what occurred or random specific memories, and make up the rest (if they are required to recall it). Dan Gilbert talks about this extensively in his book Stumbling on Happiness. Memories are not only constructive, they are reconstructive. They can be changed after the event has occurred, and not necessarily by you.

The idea of memories being altered by outside influence goes back to the 1970's with a study by Elizabeth Loftus and John Palmer (1974). They found that how the question was phrased (specifically, whether word "hit" or "smashed" was used) affected people's estimations about how fast two cars were travelling (on average, people in the "smashed" condition thought the cars were travelling 6mph faster than in the "hit" condition) and whether they would later report seeing broken glass (there was none).

So our memories can be influenced by external factors. What about internal ones? Stress has been shown repeatedly to affect delayed retrieval of a memory (e.g. Kuhlmann et al., 2005) and this has worrying implications for the validity of eye witness testimonies. Eye witness testimonies are generally seen as very powerful evidence but given how unreliable memory can be, you have to question whether it should carry so much weight. There is even a foundation set up to help those who have been incorrectly incarcerated (The Innocence Project) and about 3/4's of the convictions were based on (what were later shown to be) false memories.

Emotions are generally believed to influence memory as well, with people believing that they remember something better if it is associated with a powerful emotion. However, there is no evidence to suggest this is the case (Kensinger, 2007) and people overestimate their ability to recall emotionally charged memories.

It's not all doom and gloom though. I've painted a picture of human memory being almost worthless, when in fact it is very powerful and generally pretty good (though there are ways of improving it). But we shouldn't view our memories as infallible and we should recognise that we can incorrectly code an event into our memory or that our memory can be altered once it has been stored in our long-term memory.