Prospective Memory

Jan Rummel, Mark A. McDaniel, Jan Rummel, Mark A. McDaniel

• Publisher: Routledge
• Year: 2019
• eBook ISBN: 9781351000130
• Topic: Psychology
• Subtopic: Cognitive Psychology & Cognition
• Index: Psychology

Featuring contributions from world-leading experts, this book presents a timely overview of current theoretical, methodological, and applied issues in the field of prospective memory. The authors explore how prospective memories are formed, how they are maintained over time, and how they are retrieved. This volume integrates our understanding of prospective memory and how it functions with related cognitive processes and themes, such as context memory, metamemory, working memory, and cognitive control. Considering recent methodological advances in the field, such as the use of cognitive modeling, the book also covers individual differences in prospective memory abilities, their development across the life span, and their manifestations in naturalistic settings. The book also illustrates how the understanding of prospective memory can be integrated with other related research areas.

Prospective Memory is an invaluable resource for students and researchers of human memory.

1

INTRODUCTION

Jan Rummel & Mark A. McDaniel

When asked to think about the function of memory in everyday life, most people intuitively refer to the capacity to remember things from the past, that is, to retrospectively retrieve information that has been stored in memory earlier. However, this is by far not the only function our memory serves.

Another important memory function has been studied under the term prospective memory (referred to as PM throughout this volume). PM can be broadly defined as the mental ability to remember to perform previously formed action plans (intentions) at an appropriate moment in the future (McDaniel & Einstein, 2007). Everyday examples of PM intentions include mundane activities such as remembering to return a book to the library before it is due or remembering to take prescription pills before breakfast. However, PM also plays a pivotal role in mastering professional duties in many domains, such as in health care or air-traffic control (see Loft, Dismukes, & Grundgeiger, this volume). In light of its significance in everyday life and in many workplace settings, it is not surprising that PM has been studied extensively by cognitive psychologists as well as by applied researchers. According to the PsycINFO (1990–2018) database, the article by Einstein and McDaniel (1990) introducing a laboratory paradigm for studying PM—which soon became the standard paradigm in the field—has so far been cited 557 times in various different scientific outlets. This remarkable number of citations indicates a high research interest in PM. Also, according to the same database, the number of publications per year mentioning PM in their titles or abstracts has been continuously increasing from about 15 publications per year between 1994 and 1998 to about 30 per year between 2004 and 2008 and further to currently over 100 per year between 2014 and 2018 (see Figure 1.1).

FIGURE 1.1 The number of publications on prospective memory per 5-year interval since 1989. Data were derived from a PsycINFO database search for journal articles with the keywords (“prospective memory”) or (“prospective remembering”) or (“memory” + “intention”) occurring in the title or abstract.

Aims and scope of the present volume

As evident from the first edited volume devoted to PM, which was published in 1996 (and re-published in 2014), early PM research focused primarily on the differences between PM and other cognitive functions in order to establish PM as a new field of research (Kvavilashvili & Ellis, 1996). At that time, some memory researchers even questioned whether PM was distinguishable from retrospective memory (Crowder, 1996). Since then, however, the “zeitgeist” among memory researchers has changed; nowadays most researchers would probably agree with Ellis and Freeman’s (2008) claim from the second edited PM volume (published 12 years later) that PM is a unique ability comprising several cognitive processes and mechanisms that act together, thus allowing intentions to be remembered at the appropriate moment. Therefore, at this point PM has become a well-established field of research with a set of PM-unique theories and methods. The present chapters make it readily apparent that PM research has become richer and more nuanced, and is viewed from a growing array of theoretical perspectives.

In light of these developments, we also suggest that the time may be ripe to adopt an orientation that does not further segregate PM from other areas of cognition research, but instead better integrates PM research with theories and principles from these other areas. Because the field has been growing rapidly (see Figure 1.1), the present volume cannot provide an extensive overview of all PM research conducted since the publication of the last volume. For such overviews, we recommend the books by Cohen and Hicks (2017) and McDaniel and Einstein (2007). The present volume will instead highlight several of the broader and, in our view, most important theoretical, methodological, and applied advances that have emerged in the field during the past ten years. We hope that the reader will profit from this volume by gaining a better understanding of the complexity of the higher-order human mental capacity called PM.

To set the stage for the chapters of this volume, we will first give an overview of PM research methods. Furthermore, we will outline the terminology that has become standard in the PM literature and is used throughout this book.

What constitutes a prospective memory task and how is prospective memory assessed?

Similar to retrospective memory tasks, during which people study certain items and are asked to retrieve them after a retention interval, PM tasks are composed of an encoding and a retrieval phase separated by a retention interval. This is nicely illustrated by the so-called phase models of PM (Kliegel, Altgassen, Hering, & Rose, 2011; Kvavilashvili & Ellis, 1996), which will be used as an organizational structure in some of the upcoming chapters. These models regularly differentiate between three phases, namely (1) the intention-formation phase, during which the PM task is encoded, (2) the intention-retention phase—a time interval during which PM task execution is delayed, and (3) the intention-retrieval phase, during which the PM task has to be activated and executed at the appropriate moment.

Characterizing the appropriate moment is a critical aspect of every PM task, and two to three different types of PM tasks are distinguished in this regard, namely event-based, time-based, and sometimes also activity-based PM tasks. In event-based tasks, the intention execution is associated with the occurrence of a target event. For example, the right moment for executing the intention to share some important news with a friend may be indicated by the encounter of this friend in the hallway. The standard paradigm for studying event-based PM in the laboratory (Einstein & McDaniel, 1990) mimics such situations, while also emulating the three theoretical PM phases introduced above. First, instructions are provided for a later to-be-performed ongoing task (a short-term memory task in the early Einstein & McDaniel study or, in later studies, typically some kind of binary-choice reaction task; e.g., classifying letter strings as words or nonwords), which is then practiced. After the practice phase, participants receive instructions to form a PM intention, for example to respond with a special key to certain items (i.e., PM targets) when they occur in the ongoing task. These PM responses are typically instructed to be performed in addition to, but sometimes instead of the ongoing-task response. The following intention-retention interval is often filled with some kind of delay activity to prevent participants from rehearsing the intention. During the final intention-retrieval phase, participants perform the ongoing task they practiced before and must remember to execute the intention in response to the PM targets without being explicitly reminded of the additional PM task (Einstein & McDaniel, 1990).

Critically, PM trials are restricted to relatively infrequent occurrences to encourage participants to become absorbed in the ongoing activity, similar to people becoming absorbed in everyday activities when they also have PM tasks to remember. The concern regarding the laboratory paradigm is that if PM trials become too frequent, participants will treat the task more like a vigilance task and thus potentially mask processes that are central to PM, such as intention retrieval from long-term memory (Brandimonte, Ferrante, Feresin, & Delbello, 2001; Graf & Uttl, 2001).

For time-based tasks, intention execution is tied to a specific time of day or the expiration of a specific time period. Examples are remembering to attend a meeting at noon or remembering to take the pizza out of the oven after 12 minutes. Laboratory versions of this type of task also require participants to react in a special way at certain points in time (e.g., every 2 minutes) while performing an ongoing task (Cook, Marsh, & Hicks, 2005). Participants performing a time-based PM task are typically provided with a clock in the periphery of the ongoing task or a hidden clock that can be made visible via pressing a button while performing the ongoing task. Therefore, a unique cognitive demand imposed by time-based PM tasks is time monitoring (Waldum & McDaniel, 2016).

If the moment of intention fulfillment is indicated by a transition from one task to another, the PM task is called activity-based (e.g., remembering to call your spouse back after a business meeting). In the laboratory, activity-based PM tasks require intention execution after a specific experimental subtask has been finished (e.g., a word-categorization task that is then followed by any other kind of task; Kvavilashvili & Ellis, 1996). Activity-based and event-based PM tasks seem quite similar, but it has been argued that activity-based tasks impose fewer cognitive demands on the individual, arguably because the transition from one task to another provides a particularly good reminder of the PM task associated with it (Brewer, Marsh et al., 2011).

Some may wonder how closely these basic laboratory paradigms actually mimic everyday PM situations. Notably, other tasks that more closely resemble real-life PM situations have been developed (e.g., the virtual week task; Rendell & Craik, 2000), and many of the effects found in the basic task have been replicated with these more realistic tasks (e.g., Rose, Rendell, McDaniel, Aberle, & Kliegel, 2010).

Importantly, most basic and naturalistic PM tasks share certain features that are typically considered unique to PM (cf. Kvavilashvili & Ellis, 1996): First, the delayed intention must be executed at the appropriate moment without an external reminder. Furthermore, there are only very few instances where the intention can be executed. Finally, PM requires a person to remember that something has to be done at the appropriate moment (prospective PM component) and to remember what has to be done and when (retrospective PM component).

Measures derived from PM tasks and their meaning

PM tasks provide various measures that researchers use to draw inferences about underlying PM processes. A straight-forward measure is the number of correct responses to PM targets relative to the total number of PM target occurrences. This PM hit rate is considered the primary PM performance measure. Typically, not only immediate PM responses but also late PM responses that occur during trials directly following a PM-target presentation are considered accurate (Brewer, 2011). False PM responses (i.e., PM responses to non-target ongoing-task items) usually occur so rarely that they need not be considered. However, there are variants of the standard paradigm that foster false PM responses and use them as indicators for certain PM processes (Bugg & Streeper, this volume). Response times to PM cues have also been used as indicators of PM processing (Marsh, Hicks, & Watson, 2002), but the rare occurrence of PM targets often render these data rather unreliable.

Furthermore, it is critical in PM research to probe participants’ memory for the PM cue and the PM response after the PM experiment to ensure that participants did not retrospectively forget (or may not even have encoded) the intention. Such complete forgetting of the intention would qualify as an episodic memory rather than a prospective memory failure.

Finally, the response time and accuracy of the ongoing task are useful PM-process indicators (Marsh, Hicks, Cook, Hansen, & Pallos, 2003; Smith, 2003). Comparing ongoing-task performance in the presence versus absence of an additional PM task often reveals that the addition of a PM task imposes a cost (deceleration of responses and/or increase in errors). Whereas PM costs are a very stable empirical phenomenon, the exact processes underlying these costs are currently subject to a lively debate (see Shelton, Scullin, & Hacker, this volume; Strickland, Loft, & Heathcote, this volume). In general, PM costs are assumed to reflect the level of engagement in controlled strategic processes for the sake of the PM task. Many PM tasks seem to require such strategic PM processing and thus produce costs. However, intentions can sometimes also be spontaneously retrieved without engaging in strategic PM processing (McDaniel, Umanath, Einstein, & Waldum, 2015; see also Shelton et al., this volume).

Two pathways to successful PM fulfillment

The reliance on either type of process for PM fulfillment (i.e., strategic processing or spontaneous retrieval) is thought to depend on various factors (McDaniel & Einstein, 2000). One very important one is the quality of PM-target processing (Einstein et al., 2005). If PM targets are focal to the ongoing task, that is, if the defining features of the target are to be processed in the service of performing the ongoing task, PM can rely on spontaneous retrieval. For example, if the cue is a specific word that occurs in an ongoing lexical-decision task, a task that requires semantic word–nonword judgments, the target would be considered “focal” because its semantic content has to be processed in order to classify it as a word or a nonword. If the target is defined as a specific syllable, it is not considered “focal” within this ongoing task, because syllables are not automatically extracted when making lexical decisions (Einstein & McDaniel, 2005). Consequently, for such “nonfocal” cues, PM would need to rely primarily on strategic processing.

A high saliency of the PM target is also likely to foster sponta...

Table of contents

• Cover
• Half Title
• Series Page
• Title Page
• Copyright Page
• Table of Contents
• List of contributors
• 1. Introduction
• 2. The multiprocess framework: Historical context and the “dynamic” extension
• 3. Prospective memory in context: Methods, findings, and future directions
• 4. Fate of suspended and completed prospective memory intentions
• 5. Metacognition of prospective memory: Will I remember to remember?
• 6. Evidence accumulation modeling of event-based prospective memory
• 7. Neuropsychological and physiological correlates of prospective memory
• 8. Individual differences in prospective memory
• 9. Prospective memory across the lifespan
• 10. Take the field!: Investigating prospective memory in naturalistic and real-life settings
• 11. Prospective memory in safety-critical work contexts
• 12. Realized accomplishments in prospective memory and some thoughts about the future
• Index