How real is our memory?

Humans readily establish false memories. If you give adults a study list of words like hot, snow, warm, winter, ice, wet, chilly, weather, heat, freeze, shiver, frost, and then test them later, they will "remember" related words like cold that weren't actually on the list. They will be as sure that cold was on the original list as they are about all the words that really were there.

Small children, age 5 to 7, by contrast, are very unlikely to make this type of error. They can memorize the words on the list, but they won't generate false memories. By the time they are 11 or so, kids begin to generate false memories at about half the rate adults do. Researchers speculate this is because children haven't formed as many connections between related items and concepts, and so simply don't associate words like cold and weather as readily as adults do.

Another, related phenomenon is implicit memory. If you give adults a study list of words, then ask them to complete a word completion puzzle (What's the first word you can think of that follows this pattern: C_L_ ?), they will reply disproportionately with words on the list, and also with the related words that were not listed.

The task can be made explicit by asking participants to reply only with words from the study list. Again, they will falsely "remember" related words that they were never shown.

Kristen Diliberto-Macaluso of Berry College wanted to see if young adolescents also showed a similar pattern with implicit memory, so she conducted two experiments with fourth- and fifth-graders ("Priming and False Memories from Deese-Roediger-McDermott Lists on a Fragment Completion Tests with Children," American Journal of Psychology, 2005).

She gave the kids two types of word lists to study. One type was similar to the list shown above, where all the words were related to a "lure" word (cold, in our example). In the second type of list, the lure word was actually included in the list itself. Next she gave the kids the word completion test, asking them to respond with the first word that came to mind. Whether or not the lure was included in the list, the kids responded the same. About a third to a half the time, they answered with either a lure word or a word that was on the list, compared to around 15 percent of the time for a group of kids that had not studied the list at all.

Next, with a new group of kids the same age, she showed them the same lists of words and gave the same test, but asked that they respond only with words they had studied before. So in this case, when the lure word was not included in the list, if the response given was the lure word, then it was an example of a false memory. Over forty percent of the time, they did show false memory. However, when the lure word was included in the list, they (correctly) responded it was there about twice as often compared to the kids who had not seen the list. So, when the task was made explicit, fourth- and fifth-graders showed some false memory, but also had accurate memory for the lure words.

Overall, the false memory effect was about twice as big as the implicit memory effect for these children. The fact that the false memory effect was still smaller than accurate memory for the lure words suggests that these children's memories aren't fully mature. As we grow older, it seems, we are more able to make generalizations, but these generalizations are also more likely to lead us to error.