Findings in the laboratory do not always apply to the real-world – a myriad of factors can influence real-world phenomena, and scientists actively seek to eliminate many of them in their laboratories. But ecological validity can be particularly difficult to establish in cognitive science, where real-world levels of motivation, stress, and memory load can not always be practically (or ethically) simulated in the laboratory.
Ecological validity may be particularly important in tests of prospective memory – the ability to remember to perform a planned future action. One salient example: young children are thought to possess relatively poor prospective memory, at least for boring laboratory tasks like remembering to close a door, or to ask for a pencil. But even 2-year-olds can show remarkable prospective memory if the planned future action is something like “get ice cream!” So, clearly motivation plays a large role in prospective memory tasks, as it should: we’re fortunate that humans are particularly good at remembering to do those things deemed most important.
Laboratory tests of prospective memory (PM) are often ecologically invalid in another sense: in the real-world, if you suddenly remember something you have to do, you may not be able to execute that planned action until somewhat later. However, PM is often tested in the lab with a “retrieve-execute” paradigm, in which the planned action can be performed immediately after the PM cue is encountered. A 2006 article by Kliegel & Jager indicates that this discrepancy between laboratory and real-world PM may be significant: while the majority of retrieve-execute PM tasks show little or no effects of aging, aging effects are apparent in all studies using the more ecologically-valid “delayed-execute” version of PM tasks.
To investigate this aspect of real-world PM tasks, and how it may interact with age, working memory, and executive function, Kliegel & Jager tested 127 subjects on a delayed-execute PM task by dividing them into 4 age groups (22-31 yr olds, 60-69 yr olds, 70-79 yr olds, and 80-91 yr olds). Each subject was given a prospective memory task in which they had to monitor for the presence of certain words (e.g., “technique” or “system”) in sets of 3 sentences, each of which was followed by a two-choice comprehension question, a short pause, and then a four-choice trivia question. For half the trials, subjects had to count to 9 during the short pause (a “high “memory load condition) while for the other half of trials subjects had merely to wait (a “low” memory load condition). If subjects saw either target word, they’d have to press the space bar, but only after the pause preceeding the trivia question. In this sense, memory for the planned action was tested with a filled or unfilled delay before the plan could actually be executed. The authors also measured processing speed (via digit-symbol substitution) and inhibitory control (via Stroop interference, calculated as the difference between incongruent color words and a control condition of colored shapes).
The results showed the expected trend of slowed processing speed and decreased inhibitory control with age. Performance on the PM task (in terms of “hits”) decreased quadratically with age, and was better under low than high memory load. In the high memory load condition, the oldest adults showed decreasing performance over time, whereas the younger gropus all showed equivalent or slightly increasing performance over time. Across age-groups, processing speed and inhibitory control were both associated with better PM “hits” performance, but only inhibitory control significantly predicted PM performance after controlling for age-related effects.
The authors suggest that these results support the position active maintenance of a planned action is critically important for delayed-execution prospective memory. Although PM ability decreases with age, individual differences in the Stroop task show additional covariation with prospective memory that are not accounted for by differences in age or processing speed.
One straightforward interpretation of these results is that the Stroop task has a unique relationship with prospective memory because delayed-execution PM requires that subjects inhibit the items filling the delay. A slightly different perspective suggests that Stroop interference largely reflects the ability to actively maintain the goal of “color naming” on-line, and is related to PM insofar as delayed-execution tasks require subjects to actively maintain the goal of “respond after the pause” on-line.
In contrast to delayed-execution PM tasks, individual differences in retrieve-execute PM tasks may be more related to a capacity for maintaining vigilance and actively monitoring for the presence of the prospective memory cue than to active maintenance (to the extent that these active maintenance is dissociable from monitoring). I hope to cover this topic in future posts.
Remembering to Remember: Prospective Memory