Adult age differences in solving series problems requiring integration of new and old information

The ability of young and middle-aged adults to solve series problems was assessed with problems connected by either marked or unmarked adjectives. Half the participants given each type of problem were instructed to use imagery. Information needed to answer memory and inference questions was explicity provided, but additional information from semantic memory was needed to answer integration questions. Overall, there was no difference between age groups on the memory measure, but scores on both integration and inference measures were significantly lower for the middle-aged group, a result suggesting a deficit in the ability to manipulate items in working memory. Imagery instructions facilitated scores of both age groups equally for the integration measure, but affected neither memory nor inference scores, indicating that middle-aged adults are able to use imagery as a control process as effectively as younger adults, and also that imagery is most useful for developing an array. Finally, the only effect of adjective type was that marked adjectives significantly improved inference scores of middle-aged adults.