Growing Old or Living Long: Take Your Pick
Research to understand the psychological and emotional processes of aging is essential to creating a society in which the elderly can thrive.
The 20th century witnessed two profound changes in regions of the world where people are well educated and science and technology flourish: Life expectancy nearly doubled, and fertility rates fell dramatically. As a result, individuals and populations are aging.
Virtually all educated people are aware of the graying of the United States, yet relatively few are as aware of its implications for science, technology, and human culture. Longer life is a remarkable achievement, but now we need to apply what we are learning in the natural and social sciences to redesign human culture to accommodate long lives. We need to find cures for Alzheimer’s disease and arthritis, develop technologies that render many age-related frailties such as poor balance invisible in the way eyeglasses now compensate for presbyopia, and begin seriously rethinking cultural norms, such as the timing of education and retirement.
Longevity is the largely unexpected consequence of improvements in general living conditions. Genetically speaking, we are no smarter or heartier than our relatives were 10,000 years ago. Nonetheless, in practical terms we are more biologically fit than our great-grandparents. Robert Fogel and his colleague Dora Costa coined the term “technophysio evolution” to refer to improvements in biological functioning that are a consequence of technological advances. They point out that technologies developed mostly in the past century vastly improved the quality and sustainability of the food supply. Subsequent improvements in nutrition were so dramatic that average body size increased by 50% and life expectancy doubled. The working capacity of vital organs greatly improved. Breakthroughs in manufacturing, transportation, energy production, and communications contributed further to improvements in biological functioning. Medical technology now enables full recovery from accidents or illnesses that were previously fatal or disabling.
Even technophysio evolution may be too narrow a term. Just as dramatic as the technologies are the acceptance and incorporation of the advances into everyday life. Not only was pasteurization discovered, it was implemented in entire populations. Not only were insights into the spread of disease observed in laboratories, community-wide efforts to dispose of waste were systematically undertaken. Not only was child development better understood, child labor laws prevented little ones from working long hours in unsafe conditions. Culture changed. Life expectancy increased because we built a world that is exquisitely attuned to the needs of young people.
Remember, however, that advances of the 20th century did not aim to increase longevity or alleviate the disabling conditions of later life. Longer life was the byproduct of better conditions for the young. The challenge today is to build a world that is just as responsive to the needs of very old people as to the very young. The solutions must come from science and technology. Unlike evolution by natural selection, which operates across millennia, improvements in functioning due to technological advances can occur in a matter of years. In fact, given that the first of the 77 million Baby Boomers turned 60 in 2006, there is no time to waste. To the extent that we effectively use science and technology to compensate for human frailties at advanced ages, the conversation under way in the nation changes from one about old age to one about long life, and this is a far more interesting and more productive conversation to have.
Psychological science and longevity
In psychology, as in most of the biological and social sciences, research on aging has focused mostly on decline. And it has found it. The aging mind is slower and more prone to error when processing information. It is less adept at considering old information in novel ways. Memory suffers. In particular, working memory—the ability to keep multiple pieces of information in mind while acting on them— declines with age. The ability to inhibit extraneous information when attempting to focus attention becomes impaired. Declines are especially evident on tasks that require effortful processing that relies on attention, inhibition, working memory, prospective memory, and episodic memory.
These changes begin in a person’s 20’s and 30’s and continue at a steady rate across the adult years. They occur in virtually everyone, regardless of sex, race, or educational background. In all likelihood, these effects are accounted for by age-related changes in the efficiency of neurotransmission.
Despite these changes in cognitive processing, the subjective experience of normal aging is largely positive. By experiential and objective measures, most older people remain active and involved in families and communities. The majority of people over 90 live independently. The National Research Council report The Aging Mind: Opportunities in Cognitive Research observed that performance on laboratory tasks does not map well onto everyday functioning. The committee speculated that much of the discrepancy occurs because people spend most of their time engaged in well-practiced activities of daily routines where new learning is less critical. Research shows that in areas of expertise, age-related decline is minimal until very advanced ages.
Arguably even more interesting and important is growing evidence that performance—even on basic processes such as semantic or general memory—improves under certain conditions. One of the first such studies was reported by Paul Baltes and Reinhold Kliegl in 1992. They demonstrated rather striking improvement in memory with practice. Baltes and Kliegl first enlisted younger and older people’s participation in a study of memory training. They assessed the participants’ baseline performance and, as expected, younger participants outperformed older participants. However, after this initial assessment, participants attended a series of training sessions in which they were taught memory strategies such as mnemonics. They found that older people’s memory performance benefited from practice so much that after only a few practice sessions, older people performed as well as younger people had before they had practiced. Younger peoples’ performance also improved with training, of course, so at no point in the study did older people outperform younger people at the same point in training. But the fact that older people improved to the equivalent of untrained younger people speaks to the potential for improvement.
More recently, scientists have begun to investigate social conditions that also may affect performance. Tammy Rahhal and her colleagues reasoned that because there are widespread beliefs in the culture that memory declines with age, tests that explicitly feature memory may invoke performance deficits in older people. They compared memory performance in younger and older people under two experimental conditions. In one, the instructions stressed the fact that memory was the focus of the study. The experimenter repeatedly stated that participants were to “remember” as many statements from a list as they could and that “memory” was the key. In the second condition, experimental instructions were identical except that the instructions emphasized learning instead of memory. Participants were instructed to “learn” as many statements as they could. Once again, rather remarkable effects were observed. Age differences in memory were found when the instructions emphasized memory, but no age differences were observed in the condition that instead emphasized learning.
In another study, Thomas Hess and his colleagues documented deficits in performance when participants were reminded about declines that accompany aging before they began the experiment. In their study, participants read one of three newspaper articles before completing a memory task. One simulation reaffirmed memory decline and raised concerns that it may be worse than previously documented. In another condition, participants read a simulated article that described research findings suggesting that memory may improve with age. The third article was memory-neutral. In Hess’s study, younger people outperformed older people in all three conditions, but the gap was significantly reduced in participants who read the positive account of memory. Most important, Hess’s team identified a potential mediator of these performance differences. Participants were required to write down as many words as they could remember and those who had read the positive account about memory were more likely to use an effective memory strategy, called semantic clustering, in which similar words are grouped together. These strategic efforts were not observed in participants who were reminded of age deficits. Such findings point to the role of motivation in cognitive performance.
Thus, although there is ample evidence for cognitive deficits with age, the story about aging is not a simple story of decline. Rather, it is a qualified and more nuanced story than the one often told. Even in areas where there is decline, there is also growing evidence that performance can be improved in relatively simple ways. This poses a challenge to psychology to identify conditions where learning is well maintained, to find ways to frame information in ways best absorbed, and ultimately to improve cognitive and behavioral functioning by drawing on strengths and minimizing weaknesses.
My students, colleagues, and I had been studying age-related changes in motivation for several years. We began to wonder whether changes in motivation would affect performance on cognitive tasks, and we set out to explore what we call socioemotional selectivity theory (SST), a life-span theory of motivation.
SST was initially developed to address an apparent paradox in the aging literature. Despite losses in many areas, emotional well-being is as good if not better in older people as in their younger counterparts. Studies of specific components of emotional processing (such as physiological responses, facial expression, neural activation, and subjective feelings) suggest that this system is well maintained at older ages. Experience-sampling studies in which participants carry electronic pagers and report emotions at random times throughout their days show that negative emotions are experienced less frequently in older people and positive emotional experiences are just as frequent. Older people are more satisfied with their social relationships than are younger people, especially regarding relationships with their children and younger relatives. Fredda Blanchard-Fields and her colleagues find that older people solve heated interpersonal problems more effectively than do younger adults. Many social scientists refer to such findings as the “paradox of aging.” How could it be that aging, given inherent losses in critical capabilities, is associated with an improved sense of well-being?
Within the theoretical context of SST¸ there is no paradox. SST is distinguished from other life-span theories in that its principal focus concerns the motivational consequences of perceived time horizons. Instead of relying on the more traditional yardstick of chronological age, SST considers the effects of continually changing temporal horizons on human development. The theory maintains that two broad categories of goals shift in importance as a function of perceived time: those concerning the acquisition of knowledge and those concerning the regulation of feeling states. When time is perceived as open-ended, as it typically is in youth, people are strongly motivated to pursue information. They attempt to expand their horizons, gain knowledge, and pursue new relationships. Information is gathered relentlessly. In the face of a long and nebulous future, even information that is not immediately relevant may become so somewhere down the line.
In contrast, when time is perceived as constrained, as it typically is in later life, people are motivated to pursue emotional satisfaction. They are more likely to invest in sure things, deepen existing relationships, and savor life. Under these conditions, people are less interested in banking information and instead invest personal resources in the regulation of emotion. In this way, SST specifies the direction of the age-related motivational shift and offers hypotheses about social preferences and goals as well as the types of material that people of different ages are most likely to attend to and remember. To be clear, the theory does not speak against experience-based change. Rather, it postulates that some of the age differences long thought to reflect intractable, unidirectional change instead reflect changes in motivation. The theory thus contributes to a more nuanced interpretation of age differences.
One key tenet of SST is that perceived time horizons, not chronological age, account for age differences in goals and preferences. Our research team has examined this theoretical postulate in a variety of ways in a number of studies. We hypothesized that older people would prefer emotionally meaningful goals over informational goals but that these preferences would change systematically when time horizons were manipulated experimentally. In several studies, we showed that younger people display preferences similar to those of the old when their time horizons are shortened, and older people show preferences similar to those of the young when their time horizons are expanded. Importantly, similar changes occur when natural events, such as personal illnesses, epidemics, political upheavals, or terrorism create a sense of shortened time horizons. Under such circumstances, the preferences of the young resemble those of older people. In other words, when conditions create a sense of the fragility of life, younger as well as older people prefer to pursue emotionally meaningful experiences and goals.
Thus, when findings like those described above began to appear in the literature, my students, colleagues, and I began to apply postulates from SST to the study of age differences in cognitive processing. The human brain does not operate like a computer. It does not process all information evenly. Rather, motivation directs our attention to goal-relevant information and away from irrelevant information. We see what matters to us. Imagine walking around a city block with the goal of finding a friend. You see very different things than you would see if you took the same walk while trying to find a particular species of bird. Indeed, in the latter scenario you might walk right by your friend without notice. In the former, you would surely miss the bird.
In an initial study, my former student Helene Fung and I reasoned that because older people prefer emotional goals, they may remember emotional information better than emotionally neutral information. This was an important idea to test because the standard practice in psychological science is to avoid emotional stimuli in tests of memory in order to minimize contamination of “pure” cognitive processes. We wondered if by doing so, experimenters were inadvertently handicapping the performance of older adults. A substantial literature on memory and persuasion shows that people are more likely to remember and be persuaded by messages that are relevant to their goals. Thus, we reasoned that marketing messages that promised emotionally meaningful rewards may be more effective with older people than those that promise to increase knowledge or expand horizons.
First, we hypothesized that older people would be more likely than young people to remember advertisements that promised emotional rewards. Second, consistent with the theory, we hypothesized that modifying their time perspective would alter this preference. Recall that according to SST, age differences are due to differences in time horizons, not chronological age. Fung and I worked with a graphics design firm to develop pairs of advertisements for a range of products. The ads in each pair were identical except for the slogans. In each pair, one version had a slogan that promised an emotional reward and the other promised a future-relevant goal. For example, in a camera ad one slogan read “Capture those special moments” and the other version read “Capture the unexplored world.” In another set one slogan read “Stay healthy for the ones you love” and the matched slogan read “Stay healthy for your bright future.”
The results supported both hypotheses. In one study, older people remembered the emotional slogans and the products they touted better than did younger people, supporting our first hypothesis. To test our second hypothesis, we showed a subset of participants both versions of the ads at the same time and asked them to indicate their preference. Some were simply asked to indicate the one they liked best. Others, however, were presented with the following instruction before they were asked to indicate their preference: “Imagine that you just got a call from your physician who told you about a new medical advance that virtually insures you will live about 20 years longer than you expected and in relatively good health. Please look at these ads and tell us which one you prefer.” In this time-expanded condition, age differences were eliminated.
The positivity effect
Findings from this initial study suggested that in older people, memory of emotional information was superior to memory of other types of information. My colleagues Susan Charles and Mara Mather and I began to wonder whether such effects would be limited to emotionally positive material (most advertisements associate products with positive promises) or if there would be heightened attention to emotionally negative information as well. On this point, the theory was equivocal. Reasoning from our theory, we thought that there are (at least) two ways that emotional goals might influence older adults’ attention and memory. The first possibility was that all information relevant to emotional goals is more salient under time-limited conditions. This emotionally relevant focus would bias attention and memory in favor of both positive and negative information. The second possibility is that information that furthers emotional goals in general is more salient. This emotionally gratifying focus would bias attention and memory in favor of information that fosters positive emotional experiences and against information that generates negative emotional experiences.
A substantial literature in social psychology, albeit based exclusively on young adults, shows superior memory of negative information. Negative information is also widely believed to be weighted more heavily than positive information in impression formation and in decisionmaking. The burning question was whether such findings, long presumed to represent “human” preferences, actually represented preferences of young people.
We conducted a study in which young, middle-aged, and older adults viewed positive, negative, and neutral images on a computer screen and were then tested for their memory of the images. We found an age-related pattern in which the ratio of positive to negative material recalled increased with age. Younger people recalled equal numbers of positive and negative images. Middle-aged people showed a small but significant preference in memory for positive images. In older people, the preference for positive was striking. Older people remembered nearly twice as many positive images as negative or neutral images.
We were excited by this finding because it pointed to a particular type of information that was relatively well remembered. However, this first behavioral study did not allow us to know whether the negative images were not initially processed at all or were stored but then were less likely to be retrieved from memory. We conducted a second study using essentially the same images and procedures. In this subsequent study, we collaborated with neuroscientist John Gabrieli’s research team. We included event-related functional magnetic resonance imaging to examine brain activation while participants viewed the images. After the viewings, participants again recalled as many of the images as they could, and we computed the ratio of positive to negative images they recalled. The behavioral findings perfectly replicated those from the first study. Older adults remembered more positive images than negative images. We also observed that whereas amygdala activation increased in younger adults in response to both positive and negative images, amygdala activation was greater in older adults only in response to the positive images. These findings suggested that positive and negative stimuli are differentially encoded, pointing to attentional as well as memory processes.
At that point, we began to think that attention and memory can operate in the service of emotion regulation. That is, focusing on positive memories and images makes people feel good. Reasoning again from our motivational perspective, Mather and I posited that older people, at a subconscious or conscious level, may “disattend”to negative images. In both of the studies described above, participants had been required to look at a single image at a time. We asked whether older people, given a choice, would disengage from negative images. We designed a study in which pairs of photographs of 60 different faces were presented to participants on a computer screen. Each pair included one neutral and one emotional version of the same face. Twenty of the face pairs included a happy expression, 20 a sad expression, and 20 an angry expression. In the task, each trial consisted of the following sequence: a fixation point was displayed in the center of the screen for half a second; the neutral and emotional versions of one face were displayed in the right and left positions on the screen for one second; the faces disappeared from the screen; and a small gray dot appeared in the center of the screen location where one of the photographs had been. The dot remained on the screen until the participant pressed one of two response keys on the keyboard.
Participants were told that the study was investigating perceptual processes and that their task was to respond to a small dot displayed on the screen as quickly and accurately as possible. If they saw the dot appear on the right side of the screen, they should press the red key (the “k” key marked with a red sticker); if they saw the dot appear on the left side, they should press the blue key (the “d” key marked with a blue sticker). They were told that each time, before the dot appeared, they would see two faces on the screen and that they did not need to respond to these faces. Instead, they should just wait for the dot and respond to it as quickly as they could. Younger people responded to the dots with the same speed whether they were behind positive or negative faces. Older adults were significantly faster when the dot appeared behind the positive face than the negative face, indicating that when a neutral face was paired with a positive face, they were attending to the positive face, and that when the neutral face was paired with a negative face, they attended to the neutral face. This study provided further evidence for the favoring of positive over negative material in attentional processing in older adults.
With Quinn Kennedy, then a graduate student in my laboratory, Mara Mather and I began to reconsider the possible role of motivation in autobiographical memory. There is evidence in the literature that people remember their personal pasts more positively over time, but virtually all of the studies suffer from the inability to corroborate the accuracy of the memories. Maybe older people did have cheerier pasts than younger people. In this study, we were able to capitalize on the fact that I had collected data from an order of Catholic nuns in 1987. The project had been conducted, at the nuns’ request, to assess physical and emotional well-being in preparation for the aging of their religious community. In 1987, the mean age of the nuns was 66. In 2001, we returned and 300 of the 316 surviving nuns who had originally participated agreed to complete the questionnaires as they remembered completing them in 1987. A booklet describing media and religious events of 1987 was provided to help prime their memories for that year. Into this survey we embedded three experimental conditions. In one condition, the nuns were repeatedly encouraged to focus on their emotional states as they completed the questionnaires. In another condition, the sisters were repeatedly instructed to be as accurate as possible. Nuns in the control condition were simply asked to complete the questionnaire as they had in 1987. We then calculated the difference between 1987 and 2001 reports. Findings showed that both the oldest participants and younger participants who were focused on emotional states showed a tendency to remember the past more positively than they originally reported. In stark contrast, the youngest participants and older participants who were focused on accuracy tended to remember the past more negatively than originally reported. Among the nuns who completed the questionnaire without priming, the older nuns remembered the past more positively than they had originally reported it, and the younger nuns remembered it more negatively than they had reported it originally.
As noted above, there is substantive evidence that working memory declines with age. However, to the best of our knowledge, working memory of emotional information had never been examined empirically. Richard Davidson conceptualized affective working memory as the memory system that keeps feelings online as people engage in goal-directed behavior. Together with Joseph Mikels, Greg Larkin, and Patricia Reuter-Lorenz, we designed a study that examined working memory of positive and negative emotion in older and younger adults. Mikels and Reuter-Lorenz had previously developed a novel experimental paradigm to measure affective working memory. Images that had been normed as positive or negative were presented one at a time to participants on a computer screen. An image was presented briefly, and after a delay a second image was presented and removed. Based on memory, participants must judge which of two images was more negative (or in the case of positive trials, which was more positive). As a comparison task, participants completed a similar task that demanded judgments about the relative brightness of two images. As predicted, younger adults performed significantly better than older adults on the brightness comparison, which tested visual working memory. However, no age differences were observed in the comparison of faces, which assessed emotional working memory. Even more interesting was an interaction effect. Younger adults performed better than older adults on the negative emotion trials. But older adults outperformed younger adults on the positive emotion trials.
To summarize, whereas younger adults favor negative information as much or more than positive information, by middle age this preference appears to have shifted to a preference for positive information. Older adults show a decided preference in memory and attention for positive information. Although longitudinal studies are needed before conclusions about change over time can be drawn, cross-sectional comparisons suggest that the effect may emerge across adulthood. This “positivity effect” has been demonstrated in a range of experimental tasks that assess even the most vulnerable of aspects of cognitive processing, such as working memory. Theoretically, we argue that the pattern represents a shift in goals from those aimed at gathering information and preparing for the future to those aimed at regulating emotional experience and savoring the present.
The dark side of the positivity effect
We maintain that in general a focus on positive information benefits well-being. However, there are probably conditions when a chronic tendency to focus on positive material is maladaptive. One such context, we presumed, is decisionmaking, especially when options include both positive and negative features. When making decisions, negative features of options often have higher diagnostic value. If a person who is deciding whether to renew a health care plan remembers that she likes her physician but forgets that the plan does not pay for the hip surgery she needs, a suboptimal decision could be made.
Corinna Löckenhoff and I designed another study with two primary aims: to see whether in a decision context older people would review positive features of options more than negative features; and if this was the case, to see if we could eliminate the effect by modifying goals with instructions. Using computer-based decision scenarios, 60 older and 60 younger adults were presented with positive, negative, and neutral information about ostensible health care options. Some scenarios presented characteristics of physicians. Others presented features of health care plans. The information was hidden behind colored squares, and participants had to click on the square to see the information. They were told that positive information was behind white squares and negative information was behind black squares. We then observed how often participants examined the positive information versus the negative information. Later we tested their memory for the information. As we predicted, older adults reviewed and recalled a greater proportion of positive information than did younger adults. Most important, participants in one group were repeatedly reminded to “focus on the facts” and in this group the preference for positive information disappeared.
In one of our most recent studies, our research team with Mikels, Löckenhoff, and Sam Maglio collaborated with Stanford economist and physician Alan Garber and with Mary Goldstein, a geriatrician based at the Palo Alto Veteran’s Administration. The study examined whether older adults would make better decisions by focusing on their feelings about different options. We enlisted older and younger adults to make a series of health-related decisions. We presented the information about options under one of two instructional conditions. In one, participants were asked to focus on remembering the details about the options when making their choices. In the other, participants were asked to focus on their feelings about the options. For each decision, options were constructed so that one was clearly the better choice. Younger people performed better than older people when instructions asked them to focus on details. However, when participants were instructed to focus on their emotional reactions as they reviewed the options, the age difference was eliminated. Older peoples’ decision quality was as good as that of younger participants. Focusing on feelings when making decisions may be a good strategy for older adults.
Human need is the basis for virtually all of science. If we rise to the challenge of an aging population by systematically applying science and technology to questions that improve quality of life in adulthood and old age, longer-lived populations will inspire breakthroughs in the social, physical, and biological sciences that will improve the quality of life at all ages. Longevity science will reveal ways to improve learning from birth to advanced ages and to deter age-related slowing in cognitive processing. Longevity science will draw enormously on insights about individuals’ genomic predispositions and the environmental conditions that trigger the onset of disease, as well as identifying genetic differences in individuals who appear resilient despite bad habits. Longevity science will help us understand how stress slowly but surely affects health. Most of the challenges of longer-lived populations will require interdisciplinary collaborations. Psychological science must be a part of this process.
A complete list of the studies on SST and the positivity effect can be found at http://psychology.stanford.edu/%7Elifespan/publications.htm
P. B. Baltes and R. Kliegl, ”Further Testing of Limits of Plasticity: Negative Age Differences in a Mnemonic Skill Are Robust,” Developmental Psychology 28, no. 1 (1992): 121–125
R. Fogel and D. Costa, “A Theory of Technophysio Evolution, with Some Implications for Forecasting Population, Health Care Costs and Pension Costs,” Demography 34 (1997): 49–66.
T. M. Hess, C. Auman, S. J. Colcombe, and T. A. Rahhal, “The Impact of Stereotype Threat on Age Differences in Memory Performance,” Journals of Gerontology: Series B: Psychological Sciences & Social Sciences 58B, no. 1 (2003): 3–11.
T. A. Rahhal, S. J. Colcombe, and L. Hasher, “Instructional Manipulations and Age Differences in Memory: Now You See Them, Now You Don’t,” Psychology and Aging 16 (2001): 697–706.
T. A. Salthouse, “The Processing-Speed Theory of Adult Age Differences in Cognition,” Psychological Review 103, no. 3 (1996): 403–428.
R.T.Zacks et al.,“Human Memory,”in The Handbook of Aging and Cognition, F. I. M. Craik and T. A. Salthouse, eds. (Xxxxx, XX: Erlbaum, 2000), 293–357.
Laura L. Carstensen (firstname.lastname@example.org) is professor of psychology at Stanford University and director of the Stanford Center on Longevity. This article is adapted from the annual Henry and Bryna David Lecture, which was delivered at the National Academy of Sciences on November 14, 2006.