The researchers suggest attention is not enough to ensure accurate memory; expectation about how the information might be useful in the future is also required.
The psychologists, from Pennsylvania State University (Penn State), report their findings in the journal Cognition.
Conventional theories about memory suggest attention is the main decider of what is held in working memory. For example, the object-encoding theory says we store all the information we observe about the object we pay attention to, and the feature-encoding theory says we are more selective.
The Penn State researchers decided to test these theories by running experiments where they invited 60 participants to watch repeated scenes of a group of people playing basketball involving balls of different colors.
Object-encoding theory says the participants should remember all the information about the scene they have just witnessed - including the color of any balls. Feature-encoding theory says it is only necessary to remember the color of a specific ball if it is important to the task at hand.
The researchers invited the participants to watch videos - lasting 5-59 seconds - of scenes where actors played basketball and passed around two different colored balls.
The participants were asked to count the number of times a specific ball - the target ball - was passed between players. In each scene, after the actors began passing the target ball around, another ball of a different color was introduced and was also passed around. This was the "distractor" ball.
Most incorrect responses were the distractor ball color
Each participant watched 36 scenes, recording, after each scene, their count of the number of times the target ball was passed in that scene. The target ball color was not the same in every scene and was always different to the distractor ball. The ball colors used were red, green, blue and purple.
For the first 31 scenes, the participants were only required to record the number of times the target ball was passed. But after they were shown scene 32, an unexpected instruction appeared on the screen asking them also to record the color of the target ball in the scene they had just observed.
The results showed that 37% of the participants (22 of the 60) chose an incorrect color for the target ball, and most of these - 16 of the 22 responses - indicated the color of the distractor ball.
The team says these results show participants have memories of the color of both balls, but the memories are not attached specifically to the target ball or the distractor ball.
The researchers then ran further experiments where the participants were instructed at the outset that they had to record both the color of the target ball and the number of times it was passed in each scene.
In these experiments, only 14% of the participants got the color of the target ball wrong, compared with 37% in the experiment where they were not told in advance to record the ball color.
Expectation binds features to specific objects
One of the researchers, Brad Wyble, assistant professor of psychology, says that even if you pay attention to an object for an extended period of time it does not guarantee that you will correctly bind all the features of that object in memory.
In other words, much of what we remember is also based on our expectation of how that information might be useful in the future. Once the participants realized it was important to remember the color of the ball, their recall accuracy improved.
The researchers repeated the entire set of experiments with a different group of participants and got the same results - a finding they conclude robustly supports the idea of an expectation theory of memory, as Prof. Wyble notes:
"What we're showing is that attention is not enough to ensure accurate memory. You need some kind of expectation that attributing certain features to the object is important."
In December 2015, Medical News Today learned about a mouse study that showed how the activity of specific genes alters when new memories form, supporting the theory that the molecular basis of long-term memory relies on chemical tagging of DNA.