Decision-making, working memory, and age

In October I reported on a study that found older adults did better than younger adults on a decision-making task that reflected real-world situations more closely than most tasks used in such studies. It was concluded that, while (as previous research has shown) younger adults may do better on simple decision-making tasks, older adults have the edge when it comes to more complex scenarios. Unsurprisingly, this is where experience tells.

Last year I reported on another study, showing that poorer decisions by older adults reflected specific attributes, rather than age per se. Specifically, processing speed and memory are behind individual differences in decision-making performance. Both of these processes, of course, often get worse with age.

What these two studies suggest is that your ability to make good decisions depends a lot on whether

• you have sufficient time to process the information you need,
• your working memory is up to the job of processing all the necessary information, and
• your long-term memory is able to provide any information you need from your own experience.

One particular problem for older adults, for example, that I have discussed on many occasions, is source memory — knowing the context in which you acquired the information. This can have serious consequences for decision-making, when something or someone is remembered positively when it should not, because the original negative context has been forgotten.

But the trick to dealing with memory problems is to find compensation strategies that play to your strengths. One thing that improves with age is emotion regulation. As we get older, most of us get better at controlling our emotions, and using them in ways that make us happier. Moreover, it appears that working memory for emotional information (in contrast to other types of information) is unaffected by age. Given new research suggesting that decision-making is not simply a product of analytic reasoning processes, but also involves an affective/experiential process that may operate in parallel and be of equal importance, the question arises: would older adults be better relying on emotion (their ‘gut’) for decisions?

In Scientific American I ran across a study looking into this question. 60 younger (aged 18-30) and 60 older adults (65-85) were presented with health care choices that required them to hold in mind and consider multiple pieces of information. The choices were among pairs of health-care plans, physicians, treatments, and homecare aides. Working memory load increased across trials from one to four attributes per option. On each trial, one option had a higher proportion of positive to negative attributes. Each attribute had a positive and negative variant (e.g., “dental care is fully covered” vs “dental care is not covered”).

In the emotion-focus condition participants were asked to focus on their emotional reactions to the options and report their feelings about the options before making a choice. In the information-focus condition, participants were told to focus instead on the specific attributes and report the details about the options. There were no such instructions in the control condition.

As expected, working memory load had a significant effect on performance, but what’s interesting is the different effects in the various conditions. In the control condition, for both age groups, there was a dramatic decrease in performance when the cognitive load increased from 2 items to 4, but no difference between those in which the load was 4, 6, or 8 items. In the information-focus condition, the younger group showed a linear (but not steep) decrease in decision-making performance with each increase in load, except at the last — there was no difference between 6 and 8 items. The older group showed a dramatic drop when load was increased from 2 to 4, no difference between 4 and 6, and a slight drop when items increased to 8. In the emotion-focus condition, both groups showed the same pattern they had shown in the information-focus condition, except that, for the younger group, there was a dramatic drop when items increased to 8.

So that’s one point: that the effect of cognitive load is modified by instructional condition, and varies by age.

The other point, of course, concerns how level of performance varies. Interestingly, in the control condition, the two age groups performed at a similar level. In the information-focus condition, the slight superiority of the younger group when the load was lightest expanded significantly as soon as the number of items increased to four, and was greatest at the highest load. In the emotion-focus condition, however, the very slight superiority of the younger group at two items did not increase as the load increased, and indeed reversed when the load increased to eight.

Here’s what I think are the most interesting results of this study:

There was no significant difference in performance between the age groups when no instruction was given.

Younger adults were better off being given some instruction, but when the cognitive load was not too great (2, 4, 6 items), there was no difference for them in focusing on emotions or details. The difference — and it was a significant one — came when the load was highest. At this point, they were much better to concentrate on the details and apply their reasoning abilities.

Older adults, on the other hand, were better off, always but especially when the load was highest, in focusing on their feelings.

Performance on a digit-symbol coding task (a measure of processing speed) correlated significantly with performance in the information-focus condition for both age groups. When processing speed was taken into account, the difference between the age groups in that condition disappeared. In other words, younger adults' superior performance in the information-focus condition was entirely due to their higher processing speed. However, age differences in the emotion-focus condition were unaffected.

Younger adults performed significantly better in the information-focus condition compared to the control condition, indicating that specific instructions are helpful. However, there was no significant difference between the emotion-focus condition and the control for the older adults, suggesting perhaps that such processing is their ‘default’ approach.

The findings add weight to the idea that there is a separate working memory system for emotion-based information.

It should be noted that, somewhat unusually, the information was presented to participants sequentially rather than simultaneously. It may well be that these results do not apply to the situation in which you have all the necessary information presented to you in a document and can consider it at your leisure. On the other hand, in the real world we often amass information over time, or acquire it by listening rather than seeing it all nicely arrayed in front of us.

The findings suggest that the current emphasis on providing patients with all available information in order to make an “informed choice” may be misplaced. Many older patients may be better served by a greater emphasis on emotional information, rather than being encouraged to focus on myriad details.

But I'd like to see this experiment replicated using a simultaneous presentation. It may be that these findings should principally be taken as support for always seeking written documentation to back up spoken advice, or, if you're gathering information over time and from multiple sources, making sure you have written notes for each instance. Personally, I dislike making any decisions based solely on information given in conversation, and this is a reluctance I have found increasing steadily with age (and I'm not that old yet!).