Making mistakes is good for learning concepts regardless of age, according to a new study. Psychologists have been debating that issue for a few years, especially in light of experiments suggesting the value of trial-and-error learning declines with age—an effect that likely reflects different experimental procedures.
Much of the research psychologists do on how we learn uses a variation on the same simple paradigm: present pairs of words, such as "green tree," let people practice those pairs in some way, then present the first words in each pair and see how many of the second words people remember. In a variation, experimental subjects learn lexical, rather than conceptual, associations—instead of "green" and "tree," "qu" and "quote."
In the former case, researchers have found that eliciting mistakes actually helps people learn. For instance, asking people to guess which word related to "green" the experimenters have in mind before explaining it's "tree" reduces errors later on. In the real world, that's why pre-tests, in which teachers test students on material they haven't yet studied, can help students learn. But trial and error has the opposite effect on lexical learning.
While seniors did worse overall on lexical memory tasks, they did just as well as young people on conceptual ones.
When they dug into the details of those experiments, Andrée-Ann Cyr and Nicole Anderson found that psychologists had been giving more conceptual memory tests to younger people than older ones, which might explain why they found trial-and-error learning was good for young adults and bad for seniors. Perhaps it was the test, and not the person taking it.
To see if their hypothesis was right, Cyr and Anderson gave 32 young people and 32 older people, aged 72 on average, a conceptual memory test, and they gave another 32 young and 32 older adults a lexical test. Within each of the four groups, the researchers gave half instructions that elicited errors, first presenting one word, such as "fruit," and then asking each person to guess what fruit experimenters wanted them to learn.
Consistent with their argument, Cyr and Anderson found that trial-and-error learning—guessing and making mistakes—translated into about 10 percent fewer errors for conceptual learning and 10 percent more mistakes for lexical learning at test time. While seniors did worse overall on lexical memory tasks, they did just as well as young people on conceptual ones. An additional analysis showed that among those cases where participants had correctly recalled the answers, young and old alike remembered more of the guesses they'd made on conceptual versus lexical word pairs, suggesting that guessing worked by highlighting a set of related ideas—for example, things that are green—which reinforce memory. But precisely since the guesses one would make with a prompt like "qu" aren't already related in our minds, guessing doesn't help with the lexical tests.
"When learning emphasizes conceptual processing, error generation creates a richer memory trace" that aids recall, the authors write in the Journal of Experimental Psychology: Learning, Memory, and Cognition. "By contrast, lexical errors, in addition to [target words] are recalled significantly less, and are best forgotten in service of older adults' memory for correct information."