In prospective memory (PM) research, a common finding is that people are slower to perform an ongoing task with concurrent PM demands than to perform the same task alone. This slowing, referred to as costs, has been seen as reflecting the processes underlying successful PM. Historically, costs have been interpreted as evidence that attentional capacity is being devoted toward detecting PM targets and maintaining the intention in working memory; in other words, the claim is that participants are monitoring. A new account, termed delay theory, instead suggests that costs indicate a strategic speed/accuracy adjustment in favor of accuracy, allowing more time for PM-related information to reach its own threshold. Taking a meta-analytic approach, we first review studies in the PM literature that have reported ongoing task performance, both with and without a concurrent PM task, identifying key factors suitable for the meta-analysis. Next, we analyze the data of these studies, using our factors as moderators in a series of metaregressions, to determine their impact on the presence or magnitude of PM-related costs. Finally, we interpret the results of the meta-analysis from both monitoring and delay perspectives in an effort to better understand the nature of costs and what they reflect about the underlying cognitive processes involved in PM.

The goal of this research was to determine whether and how people deactivate prospective memory (PM) intentions after they have been completed. One view proposes that PM intentions can be deactivated after completion, such that they no longer come to mind and interfere with current tasks. Another view is that now irrelevant completed PM intentions exhibit persisting activation, and continue to be retrieved. In Experiment 1, participants were given a PM intention embedded within the ongoing task during Phase 1, after which participants were told either that the PM task had been completed or suspended until later. During Phase 2, participants were instructed to perform only the ongoing task and were periodically prompted to report their thoughts. Critically, the PM targets from Phase 1 reappeared in Phase 2. All of our measures, including thoughts reported about the PM task, supported the existence of persisting activation. In Experiment 2, we varied conditions that were expected to mitigate persisting activation. Despite our best attempts to promote deactivation, we found evidence for the persistence of spontaneous retrieval in all groups after intentions were completed. The theoretical and practical implications of this potential dark side to spontaneous retrieval are discussed.

Relatively little research has focused on how prospective memory (PM) operates outside of the laboratory, partially due to the methodological problems presented by naturalistic memory research in general and by the unique challenges of PM in particular. Experience sampling methods (ESM) offer a fruitful avenue for this type of research, as recent work from Gardner and Ascoli (Psychology and Aging, 30, 209-219, 2015) has shown. They found that people thought about PM around 15% of the time, and that future thinking was more common than past thinking. In two studies, we replicated our own findings and those reported by Gardner and Ascoli. To summarize, people think about the future more often than the past (30% compared to 13%), and PM occupies our thoughts approximately 13-15% of the time, supporting claims made by some researchers that our episodic memory systems are forward-looking (Klein in Journal of Applied Research in Memory and Cognition, 2, 222-234, 2013). Of those PM thoughts, participants reported that 61% were internally cued, rather than externally triggered. Through the use of multi-level modeling, we additionally found that PM thoughts were more likely when the respondant was alone than with people, and earlier in the day. Finally, we found that participants higher in neuroticism were more likely to report thinking of PM, and that this was driven entirely by the anxiety facet. Most generally, we hope to have demonstrated the value of ESM to help researchers investigate and understand naturalistic PM.

In prospective memory (PM) research, costs (slowed responding to the ongoing task when a PM task is present relative to when it is not) have typically been interpreted as implicating an attentionally demanding monitoring process. To inform this interpretation, Heathcote, Loft, and Remington (2015), using an accumulator model, found that PM-related costs were associated with changes in a decision threshold parameter. This pattern was interpreted as disfavoring a monitoring process and supporting a non-capacity-consuming delayed responding strategy. The present study combined both behavioral and modeling techniques, as well as embedded parameter validation, to better illuminate the underlying processes involved in PM. We encouraged participants to use either a delayed responding or a monitoring strategy and used these conditions as anchor points for comparing a standard PM condition (with no strategy instructions). The monitoring strategy benefited PM more than did a delayed responding strategy. Most importantly, behaviors and modeling parameters associated with the standard PM instructions more closely reflected footprints of monitoring. Further, we found no individual model parameter that directly implicates monitoring behavior. (PsycINFO Database Record

No abstract