[Differential effects of aging on processes involved in mental imagery generation]

The purpose of this study was to measure the degree to which the deterioration of mental image generation processes observed in elderly subjects by Dror and Kosslyn (1994) depends on the nature of the processing (referential vs. representational) involved in performing the task at hand. In keeping with the results of research showing that processing resources diminish as normal aging occurs (Craik & Bird, 1982; Rabinowitz, Craik & Ackerman, 1982), we suggest that generation ability based on referential processing is affected to a greater extent in older adults than is the generation ability involving representational processing. Fifteen young adults and 15 elderly persons performed a generation task modeled on that used by Dror and Kosslyn (1994). Observations were based on two contrasting conditions. In the first instance, subjects were required to generate an image representation of numbers when instructed to do so by a graphic stimulus, e.g., "31" (representational generation). In the second case, generation was produced on the basis of a verbal stimulus, e.g., "thirteen" (referential generation). Results (response time and number of accurate answers) show that, compared to young adults, elderly adults have a diminished ability to generate images based on words rather than numbers and do so at a slower pace. Our results suggest, therefore, that the performance of the first type of generation is cognitively more demanding for elderly subjects than that which is founded on figurative stimuli. The deterioration of mental image generation processes observed in elderly subjects by Dror and Kosslyn (1994) is therefore dependent on the nature of processing (referential vs. In conclusion we suggest two complementary assumptions to explain the observed dissociation. The differential effect of age may be the result of a deficienness cy linked to the nature of processing (control vs. automatic). Our second assumption is that, more specifically, referential generation involves more processing resources than does representational generation because it requires an additional transcoding operation. In this instance, the differential effect of aging that we observed would be closely linked to the number of cognitive operations involved in mental image generation.