Strategic learning of people's names as a function of expected utility in young and older adults.

People's names are challenging to learn at all ages. Because people somewhat know this, they might spontaneously use cost-efficient encoding strategies and devote more resources to learn names that are most likely to be useful. To test this hypothesis, we created a pseudo-incidental learning situation in which young and older participants were exposed to 12 characters from a TV show and reviewed face-name-instrument triplets. Characters' probability of appearance was specified via importance labels (main or secondary characters, bit parts). A surprised cued recall test showed that young adults performed better than older ones, and that semantic information was better recalled than names. Consistent with cost-efficient encoding strategies, participants in both groups recalled names and semantic information about most important characters better. Interestingly, there were large individual differences: people who reported using cost-efficient strategies performed better. At the individual level, memory advantages for most important characters' names and semantic information correlated.