Eyewitnesses: how accurately do people recall events?

As a writer you rely, directly or indirectly, on recollection of events.  For many years, psychologists have been studied how well (or badly) people can remember experiences that they witness.  

In particular, the question of whether an eyewitness can give an accurate report in a courtroom has been studied.  In Europe at the start of the 20th century there was even a journal devoted to the issue.  One of the best known contributors was Hugo Münsterberg, a German Psychologist who recounts an experiment conducted by a colleague, in which one student apparently shot another in a lecture theatre.  In fact the incident was staged, and the witnesses were given a memory test afterwards.  The smallest number of mistakes stood at 26%, and the largest was 80%.

Münsterberg's work was devoted to the many applications of Psychology to everyday life, including business, education, medicine and law.  In his 1908 book 'On the Witness Stand' he begins by listing numerous errors in his own recollection of a burglary.  His report to the police turned out to be wrong in many respects, and he found it hard to know why:

"How did all those mistakes occur? I have no right to excuse myself on the plea of a bad memory. During the last eighteen years I have delivered about three thousand university lectures. For those three thousand coherent addresses I had not once a single written or printed line or any notes whatever on the platform; and yet there has never been a moment when I have had to stop for a name or for the connection of the thought..."

Münsterberg, who moved from Germany to the USA at the invitation of William James, eminent psychologist and brother of novelist Henry James, believed that the errors originate in the fast and inaccurate associations we make as we perceive the event: "The sources of error begin, of course, before the recollection sets in. The observation itself may be defective and illusory; wrong associations... may make it imperfect."  This puts me in mind of Blink - Gladwell's (2005) book on the fast, largely unconscious parts of processing which are highly adaptive - often quicker and better than conscious thought, but which sometimes, due to stress or stereotypes, can get things very wrong!

Much better known nowadays is the work of American psychologist Elizabeth Loftus.  Loftus has shown that misremembering of eyewitness accounts is commonplace, and as well as our stereotypes, it can also be influence by information received after the event - for example, police questioning.  Amazingly, the work of Loftus and colleagues has even shown that completely fictional 'memories' can be created by repeated questioning (Ceci et al, 1994).

All of which show that what we remember, or what other people who we interview remember, might be a lot less reliable than we like to think.

References
Ceci, S.J., Loftus, E.J., Leichman, M.D. and Bruck, M. (1994). The possible role of source misattribution in the creation of false beliefs among preschoolers. International Journal of Clinical & Experimental Hypnosis, 42, 304-320.
Gladwell, M. (2005).  Blink: The Power of Thinking Without Thinking.  New York: Little, Brown.
Münsterberg, H. (1908).  On the Witness Stand:  Essays on Psychology and Crime.  New York: Doubleday.