Memory distortions: Not just for long-term memories

Consider the following list of words:

Spring
Egg
Basket
Jellybean
April
Hunt
Ham
Sunday
Chocolate
Peeps
Bonnet

If you had to memorize this list for a test, you’d actually do a pretty good job.  Why?  Well, you wouldn’t have to remember each word individually—you could just remember the overall gist of Easter.  Simply remembering the gist would help you generate the words since you’d already know that the words go together.

We often remember things by relying on the overall gist of an event—for example, instead of storing every detail about our last birthday, we tend to remember abstract things like “I had a fun party” or “I was in a grumpy mood because I felt old.”  This strategy allows us to remember more things about an event, but there’s one major drawback: by storing memories based on gist, we actually change how we remember the event.  This happens because we are biased to remember things that are consistent with our overall summary of the event.  So if we remember the birthday party was “super fun” overall, we’ll exaggerate how we remember the details—the average chocolate cake is now “insanely good”, and the 10 friends who were there becomes a “huge crowd.”  One of the factors that could contribute to this distortion is time; as you forget the details of an event, there’s more room for gist to change how you remember things.  But you would remember the details of an event immediately afterward, right?

As it turns out, gist changes the way we remember an event after just one second.  In a recent study, Brady & Alvarez (2011) asked participants to remember the sizes of red and blue circles that varied in size.  Sometimes, the blue circles were larger on average than the red circles, and sometimes the red ones were larger.  After studying the red and blue circles briefly (but ignoring the green ones), the participants waited one second and were then tested on their memory for the size of one of the circles.

A single, black circle appeared at one of the colored circles’ locations, and participants had to resize this circle to match the circle that had previously occupied that location (the bottom left location in the displays above).  When the circle at the tested location had been viewed amongst the larger circles (as depicted in the right display), participants remembered the test circle as being bigger than it actually was.  But when the same circle had been seen in the context of smaller circles (see left display), participants remembered the test circle as being smaller.

Thus, even our short-term memory depends on the overall gist of an event, not just the individual pieces.  Like long-term memory, this strategy could allow us to store more details about an event in the short-term.  But it also means that our short-term memories are subject to the same distortions as long-term memory.  This isn’t a big deal if we’re just studying circles—but what if we have to pick out a perpetrator from a lineup?  We may not be able to maintain a precise image of the real perpetrator if our short-term memory is contaminated by other people in the lineup.

And lest you think that the participants in the study were just bad at remembering things in general, go ahead and test yourself on the Easter word list.  How many words did you remember?

More importantly, was this guy on your list?

Brady TF, & Alvarez GA (2011). Hierarchical encoding in visual working memory: ensemble statistics bias memory for individual items. Psychological science : a journal of the American Psychological Society / APS, 22 (3), 384-92 PMID: 21296808