Positivity Effect: Why Older Adults Remember More of the Good

The leading account is motivational. Socioemotional selectivity theory, set out by Carstensen, Isaacowitz and Charles in 1999, holds that people's goals shift with their sense of how much time is left. When the horizon feels open, goals about gathering information and preparing for the future dominate, and negative information is useful. When the horizon feels limited, goals about emotional meaning and well-being move to the front, and attention and memory are spent accordingly. On this account the positivity effect is not a decline in anything. It is what emotion regulation looks like when it is applied to what gets noticed and kept.

Several lines of evidence favor the motivational reading over the alternative that older brains simply process negative material less well. The effect is larger when participants are free to attend as they like than when a task forces them to process every item; Reed, Chan and Mikels found in their 2014 meta-analysis of 100 studies with 7,129 participants an overall positivity effect of about a quarter of a standard deviation, rising to nearly half a standard deviation in studies that did not constrain processing. Within age groups, older adults showed a small positive bias and younger adults a small negative one. The effect also moves with the time horizon rather than with age itself: Barber, Opitz, Martins, Sakaki and Mather showed in 2016 that asking younger adults to imagine a limited future made their recall more positive, in the direction older adults show without instruction. And it can be switched off: when older adults are asked to focus on accuracy or on gathering facts, as in Löckenhoff and Carstensen's 2007 health-decision study, the age difference shrinks.

Three limits keep the effect in proportion. It is a difference in relative emphasis, not an inability to remember bad things; older adults in the 2003 study still recalled negative pictures. It depends on having cognitive resources to spare, so it is weaker under load or distraction, which fits the idea that it takes effort to regulate attention. And it is a group average with wide individual variation, which is why the demo on this page tested a prediction about groups rather than claiming to read the effect off any one person's two memory checks.

Susan Turk Charles, Mara Mather and Laura Carstensen published "Aging and emotional memory: The forgettable nature of negative images for older adults" in the Journal of Experimental Psychology: General in 2003. In the first study, 144 adults in three age groups, younger adults in their twenties, middle-aged adults in their forties and fifties, and older adults aged 65 and over, viewed 32 pictures on a computer screen, 27 of them from the International Affective Picture System and five neutral images of people added because that set had too few: 8 positive, 8 negative and 16 neutral. After a distraction task they first recalled as many pictures as they could, describing each, and then picked the pictures they had seen out of a set of old and new ones.

Recall fell with age overall, as it does for most material. The finding the paper is known for is the change in the mix. The relative number of negative pictures recalled, compared with positive and neutral ones, decreased with each successively older age group, and recognition memory showed a similar decrease with age in the relative advantage for negative pictures. A second study found that the largest age differences in both recall and recognition accuracy were for the negative pictures. The authors read both results through socioemotional selectivity theory: older adults invest more in regulating emotion, and the negative pictures are what that investment leaves behind.

The 2003 paper sits in a larger program. Mather and Carstensen's 2005 review in Trends in Cognitive Sciences collected parallel findings in attention, where older adults attended less to negative faces and more to positive ones than younger adults did, and in autobiographical memory, where older adults remembered their own past choices and feelings more positively. Reed, Chan and Mikels's 2014 meta-analysis then tested the whole literature and found the effect reliable and moderated in the ways the theory predicts: larger when processing is unconstrained and when the age gap between groups is wider.

The rows below are illustrations, not findings from the studies cited on this page.

The effect mostly does not need countering; it is part of how emotional well-being is maintained in later life, and the stereotype it overturns does more harm than the effect does. It needs attention where a decision depends on the negative information, and there the research itself supplies the method.

• Change the goal, not the person. Löckenhoff and Carstensen found that asking older adults to focus on the facts of a health decision reduced the age difference in attention to negative information. A request to review the downsides works better than an assumption that they were missed.
• Put the negatives in writing. Attention is where the effect operates, so a written list of risks and costs, read back, survives the selective summary.
• Separate understanding from emphasis. A person who remembers the good parts of a diagnosis has usually understood the rest; ask before assuming.
• For younger readers, notice the mirror image. The meta-analysis found a small negative bias in younger adults, so the correction runs the other way: a bad week is weighted more heavily than the same week will be weighted later.
• Drop the stereotype. The research does not describe older adults as naive or in denial; it describes a shift in what is attended to when time feels short, one that younger adults show too when they imagine a limited future.

The positivity effect is a difference between age groups in what gets attended to and kept. Its neighbors are biases within a person.

• It is not the reverse of negativity bias, the greater weight of negative information in attention, memory and judgment. Negativity bias is the general pattern, and the small negative bias of younger adults in the meta-analysis is an instance of it; the positivity effect is the finding that older adults do not show it, and lean the other way.
• It is not rosy retrospection, rating a past experience more positively than it was rated at the time. That happens at every age; the positivity effect is the age difference, and it is measured mostly in attention and in memory for material seen minutes earlier.
• It is not the fading affect bias, in which the emotion attached to unpleasant memories fades faster than the emotion attached to pleasant ones. That is a change in feeling over months within one person; the positivity effect is a difference in what is encoded and recalled.
• It is not stereotyping of older adults, although it belongs with it on this site, because the stereotype of the gloomy elder is the generalization the research contradicts.