Biased retellings of events yield biased memories

When we tell people about things that have happened to us, we shape the stories to our audience and our purpose. The amount of detail we give and the slant we give to it depends on our perceptions of our audience and what we think they want to hear. Does this change our memory for the event? Certainly we are all familiar with the confusion we get after we have been telling a particular story for years — we’re no longer sure what really happened and what we’ve added or subtracted to make a better story.

There is a popular misconception that dramatic public events such as earthquakes, the Challenger disaster, JFK’s assassination, etc, are given a special status in memory (“flashbulb memories”). It is true that people often have very clear memories of these events, but research has shown that such memories are as likely as any other event memory to be inaccurate1. Vividness, I regret to say, is no reliable measure of the accuracy of a memory.

This study used two invented stories rich in detail that could be told from more than one point of view. After studying one or other of the stories, subjects were asked to write a letter about one of the characters, the letter to be biased either for or against the character. Control subjects were simply asked to write as much as they could remember about the specified character. Subjects were later tested on their recall of the original story. Different aspects of memory were investigated in a series of four experiments.

It was found that subjects who wrote the biased letters recalled more information about the specified character that was related to the biased perspective. Their recall of the other character in the story was unaffected. They also added more (erroneous) details about the character — these errors being consistent with the particular bias they’d been given.

Although selective rehearsal (the fact that these subjects had had an opportunity to rehearse the information that supported the appropriate bias, at the expense of other information) plays a part in this, biased memory was found even when the subjects, in the fourth experiment, were asked to write a biased evaluation instead of a biased retelling (to avoid rehearsal of specific items).

The slant we give to information guides our encoding of the memory, the way we organize it, and the connections we make to other memories.


1. Neisser, U. & Harsch, N. 1992. Phantom flashbulbs: False recollections of hearing the news about Challenger. In E. Winograd & U. Neisser (eds.) Affect and accuracy in recall: Studies of 'Flashbulb Memories'. New York: Cambridge University Press.

Larsen, S.F. 1992. Potential flashbulbs: Memories of ordinary news as the baseline. In E. Winograd & U. Neisser (eds.) Affect and accuracy in recall: Studies of 'Flashbulb Memories'. New York: Cambridge University Press.

  • The vividness and clarity of a memory is no guarantee of its accuracy
  • Memories of dramatic public events have no particular status in memory, and are as likely as any other memory to be inaccurate
  • The retelling of events changes our memory of them
  • If we perceive an event from a particular biased perspective, or adopt a biased perspective when retelling the event, it will distort our memory of the event accordingly

Tversky, Barbara & Marsh, Elizabeth J. 2000. Biased retellings of events yield biased memories. Cognitive Psychology, 40, 1-38.