What processes allow us to execute delayed intentions? This ability, known as prospective memory, is often considered to have two constituent parts: a prospective component which involves forming the intention and possibly maintaining it until action execution, and a retrospective component which involves retrieving this intention, if that intention is not successfully and continuously maintained until the moment of action execution.
These components can be easily illustrated. Imagine yourself in a situation where prospective memory is required: while at work, you realize that you need to pick up more coffee beans. Depending on how long you must wait to acquire more coffee, you may think about it continuously until you pick it up (a very costly strategy, in terms of attention and mental effort). Thus far we have only dealt with the prospective component - forming an intention and perhaps maintaining it continuously. On the other hand, you may lapse in maintaining the intention until you are actually driving by the coffee shop. This is a less costly but more risky strategy - what if you are deeply distracted at the moment you drive by? This is the retrospective component, and seems to involve both noticing a the "cue" and retrieving the associated intention.
Lifespan trends in prospective memory may be particularly pronounced. Older adults are notorious for their inability to remember to do certain things - in other words, for prospective memory failures. Likewise, children are sometimes said to have little regard for the future consequences of their actions - does this also reflect some kind of failure of prospective processes?
One challenge to this idea is the recent finding that children, even preschoolers, can indeed complete prospective memory tasks. But several aspects of this work by Kliegel & Jager indicate that children may be completing these prospective tasks with purely retrospective processing.
To support this claim, I have to detail some of Kliegel & Jager's methods/results. These are presented in italics below, followed by my interpretations.
The authors gave 119 2, 3, 4, 5 and 6 year-olds an event-based prospective task in which children had to name the object depicted on each of 30 cards, but had to remember to place any card depicting an apple into a box. For some children, the authors placed the box in a visible location, along with an apple, as an external reminder of both their delayed intention (placing apple cards in the box) and the task's prospective cue (apples). To examine the influence of this continuously-visible reminder, other children had the box placed behind them. In addition, all parents completed a questionnaire evaluating children's prospective and retrospective memory failures in everyday life called the PRMQ. This questionnaire contains subscales which seem to reflect distinct retrospective and prospective components (as demonstrated by factor analysis on data from adults aged 17 to 94).
The results showed that 2 and 3 year olds were significantly worse at executing the prospective memory than 4, 5 and 6 year-olds, but that there were no further significant age differences. In addition, children who had a continuously visible reminder performed better on average. All of these effects remained when controlling for ongoing task performance (indicating that these age differences aren't just a result of differences in object naming abilities). Finally, when children were asked to say what they should have done with apple cards at the end of the task, the vast majority of 2 year olds (but not other age groups) failed to remember the instructions.
The fact that most 3-year-olds successfully remembered the instructions is somewhat bizarre, given their otherwise poor performance at actually executing those actions. In addition, the continuously-visible reminder of the intention was particularly helpful for that age group's performance, relative to other age groups. Therefore, one might speculate that three-year-olds have difficulty with this prospective memory task because they fail to retrieve the delayed intention at the appropriate moment. When asked or cued with a continuously available reminder, they can perform well, but otherwise they seem deficient in using the prospective memory cue to retrieving and execute the intention.
This hypothesis gains support from the relationship of the PRMQ to task performance. Multiple regression demonstrated that retrospective but not prospective components were predictive of performance on this task, even when controlling for the effects of age, and even when the data was limited to just those who were able to successfully remember the instructions when asked at the end of the task. In other words, individual differences in prospective memory performance among preschoolers seems to be influenced by age and also retrospective (but not prospective) ability.
The potential conclusion is not lost on the authors, who remark that "the present findings might reflect the initial evidence that prospective and retrospective memory abilities may be closely related during the early stages of cognitive development" and that "the present study seems to add further evidence to this proposal by revealing a high proportion of 2-year-olds who were unable to perform the intended action, presumably because of poor retrospective memory for the prospective memory task instructions."
In general, these results support the claim that prospective memory performance is primarily retrospective among very young children. A first level of competence may be the ability to retrieve the prospective instructions when cued, as reflected by 2-year-olds difficulty with recalling the instructions at the end of the task. A second level of competence may involve recognizing prospective memory cues in the context of an ongoing task, as reflected by the helpfulness of continuously-visible reminders for 3-year-olds. A third level of competence may involve successfully retrieving the cued intention, as reflected by the predictive value of retrospective but not prospective everyday memory failures on age-controlled individual differences in this task.
The implication of this work is that retrospective components of prospective memory may be relatively uncontaminated by prospective components in this age group. This provides a unique opportunity to study retrospective components of prospective memory in isolation, which may not be possible in studying other age groups. Furthermore, it implicitly challenges other accounts of cognition in preschoolers which posit an important role for prospective processes.