Memory Training

Research on the effectiveness of memory training in older adults is important for practical and theoretical reasons. Much of the applied research on this topic is motivated by the aim of trying to identify the techniques that are most effective for improving the memory of older adults. Memory training research also has importance for theories of aging because knowing the potentials and limits of memory function bears on questions of cognitive plasticity in adulthood and aging. This entry summarizes the effectiveness of memory training studies with older adults and reviews some of the techniques used for trying to improve memory function.

Memory functioning decreases with advancing age. Across many studies, the age difference between younger and older adults in recall and recognition from episodic memory is about one standard deviation (see La Voie and Light; Verhaeghen, et al., 1993). The negative effects of aging on memory must depend in large part on age-related declines in the efficiency of brain mechanisms. For example, neuro-imaging studies have provided descriptions of age-related differences in the recruitment and activation of the brain mechanisms that are involved in memory function (for a review, see Raz). However, memory performance depends not only on neurobiological factors but also on strategies for remembering and retrieving information. Training studies demonstrate that some portion of the memory decline of older adults has to do with the use of nonoptimal learning and memory strategies, and that older adults can learn to use more effective strategies so as to improve memory.

Programs designed to improve the memory of older adults generally do produce improved memory performance. In a meta-analysis of thirty-two studies, based on data from 1,539 persons, Verhaeghen et al. (1992) reported that memory training boosted performance by 0.73 standard deviation. The effects of training on performance were larger than the effects of mere retesting (0.38 standard deviation) or placebo treatments (0.37 standard deviation). The results of the meta-analysis also showed that the effects of memory training appear to be durable over six-month periods or longer after training. Across studies, performance gains associated with training were larger when participants were told in advance about the nature of the training. Gains were also larger when training was carried out in groups rather than individually, and when training sessions were relatively short.

Memory training programs produce improvements in the person's subjective evaluations of memory as well as in actual memory performance. Subjective evaluations of memory functioning are typically measured using self-report questionnaires. In a meta-analysis of twenty-five studies that examined the effects of memory training on subjective measures of performance, Floyd and Scogin reported that the magnitude of improvement is less for subjective measures than for objective measures. Subjective evaluations of memory performance improved by about 0.2 standard deviation as a result of memory training. As with the results for objective measures, subjective measures were enhanced by including pretraining information about the use of memory skills such as imagery.

Comparing across studies in the Verhaeghen et al. meta-analysis, the pattern of effect sizes for objective measures of improvement indicated that no one type of training procedure was any more effective than any other. In a study directly comparing the effectiveness of several types of memory training procedures, Rasmusson et al. reported that there was no evidence to suggest that any one type of training was superior to any of the others used. In this study, residents of a retirement community age sixty-five to ninety-two years of age were given a microcomputer-based memory training program, a commercially available audiotape memory improvement program, or a group-based memory course in weekly ninety-minute sessions for nine weeks.

Retrieval of information from memory is likely to be better when information is distinctly encoded and systematically organized or stored or filed. Frequently the procedures used in memory training studies are variations on a method that teaches individuals to associate items to be remembered with a familiar series of locations. In this method, called the method of loci, individuals are taught to remember lists of items by forming visual associations between the nth item in a list and the nth place or locus within a familiar sequence of loci. Retrieval of the items occurs by mentally traveling through the familiar sequence, and retrieving the associated item at each locus. Some writers have noted that ancient Roman orators used this procedure as an aid for remembering the main points or themes in long speeches. The orators would first memorize a large number of places in a serial order, such that each locus could be clearly visualized. Next, after a speech was prepared, its content was divided into a series of visual images which represented key words or themes in the speech. Each of these images was serially associated with one of the loci. For example, the first theme in the speech would be visually associated with the entrance to a building; the second idea would be associated with the second place in the building; and so on. To recall the main themes of the speech, one simply imagined traveling through each of the places in the building.

A similar mnemonic technique is the peg word method. Images of concrete objects rather than locations are used as the pegs to which the images to be remembered are attached. This method requires the person to readily retrieve both the peg words and their order. In a rhyming peg word method, for example, each peg word rhymes with the number indicating its position in the list: "One is a bun, two is a shoe, three is a tree," and so on.

It is important to emphasize that improvements associated with memory training effects are specific to the type of training provided. That is, there is little or no evidence to suggest that general-purpose memory function can be improved by training. In other words, memory training probably does not affect general speed of processing or brain plasticity per se, but instead exerts its beneficial effects by instilling specific strategies for the effective retrieval of specific kinds of information. There is very little if any work on the training of working memory or on the training of speed of processing that seems to underlie most, if not all, age-related deficits in cognitive performance (e.g., Salthouse). Measures of the extent to which the fundamental processes of memory could be improved with training might provide a metric for the description of individual differences in the potentials and limits of cognitive plasticity. Along these lines, aging is associated with reduced cognitive plasticity or cognitive adaptability, and training or testing the limits generally serves to enlarge or amplify the magnitude of age differences (e.g., Baltes; Baltes and Kliegl; Kliegl et al.; Verhaeghen and Marcoen).

It is important to point out that the descriptions of the negative effects of aging on memory are drawn almost entirely from studies using measures of recall or recognition of previous events or ideas (for a review, see Zacks et al.). However, many of the situations that require memory in everyday life involve keeping track of things to do or remembering to do something in the future (e.g., remembering an appointment). Only a few studies have examined the remediation of age-related declines in prospective memory (see Villa and Abeles). That prospective memory tasks can easily be relegated to external aids such as lists and calendar notations suggests that effective memory functioning involves distinguishing what information is best to commit to memory and what information is best to assign to external aids. Perhaps memory intervention programs could be designed to be even more effective or useful by giving training on what to remember as well as on how to remember.