Revealing the dynamics of prospective memory processes in children with eye movements

https://doi.org/10.1016/j.ijpsycho.2020.12.005Get rights and content

Highlights

  • Eye tracking reveals prospective memory (PM) processes in 6–7 and 9–10 year olds.

  • Children do not show monitoring behaviour during a scene viewing task.

  • Children search for the intention after spontaneously noticing the PM cue.

  • Older children's processes starting with fixating the PM cue are faster.

  • Working memory influences processes starting with the first fixation on the PM cue.

Abstract

Prospective memory (PM), the memory for delayed intentions, develops during childhood. The current study examined PM processes, such as monitoring, PM cue identification and intention retrieval with particular focus on their temporal dynamics and interrelations during successful and unsuccessful PM performance. We analysed eye movements of 6–7 and 9–10 year olds during the inspection of movie stills while they completed one of three different tasks: scene viewing followed by a snippet allocation task, a PM task and a visual search task. We also tested children's executive functions of inhibition, flexibility and working memory. We found that older children outperformed younger children in all tasks but neither age group showed variations in monitoring behaviour during the course of the PM task. In fact, neither age group monitored. According to our data, initial processes necessary for PM success take place during the first fixation on the PM cue. In PM hit trials we found prolonged fixations after the first fixation on the PM cue, and older children showed a greater efficiency in PM processes following this first PM cue fixation. Regarding executive functions, only working memory had a significant effect on children's PM performance. Across both age groups children with better working memory scores needed less time to react to the PM cue. Our data support the notion that children rely on spontaneous processes to notice the PM cue, followed by a resource intensive search for the intended action.

Introduction

Prospective memory (PM) describes the ability to perform a delayed intention at a specific event or time in the future. PM is highly prevalent from early childhood on (Winograd, 1988) and is important for an independent life (Altgassen et al., 2017). A good example of an everyday PM task in childhood might involve remembering to wash hands when coming home. Somerville et al. (1983) found that children with 2 years of age were able to remind their mothers to buy candy in a store, while Kliegel and Jäger (2007) reported a reliable ability to solve PM tasks only at the age of 3 years. In middle childhood the importance of PM increases as children need to remember to do homework and give parents messages from their teachers. It seems widely accepted that PM performance increases during childhood (for a review of studies which support or do not support this claim see Kvavilashvili et al., 2008). However, it is still under discussion, which processes underlie PM, how these processes develop and whether these processes change during the course of a PM task. The current study will contribute to the ongoing discussion by examining the processes that are involved in successfully performing a PM task. In particular, the temporal dynamics of these processes within a single trial as well as during the course of a PM task will be investigated. We therefore measured eye movements of children of two age groups (6–7 and 9–10 years of age). As to our knowledge this is the first study researching PM in children using eye tracking.

Performing a PM task successfully requires an interplay of multiple processes. It starts with forming an intention, such as “I will read my textbook and whenever seeing an exercise symbol I will answer the study questions”. Then, the intention and the intended action have to be encoded into memory, followed by a retention interval before the performance interval can start (Ellis, 1996). During the performance interval (i.e. reading the textbook) the appropriate moment to implement the intended action might occur. This performance interval consists of the so-called ongoing task (i.e. reading) and the event signalling the appropriate moment to act (the PM cue; i.e. the exercise symbol in our example) could occur. The processes of the performance interval have been widely debated (e.g., Einstein and McDaniel, 2010; Smith, 2010). It is assumed that successful PM performance either requires preparatory attention processes like monitoring for the PM cue (Smith, 2003; Smith and Bayen, 2004), spontaneous processes to notice the PM cue (McDaniel and Einstein, 2000) or a temporally dynamic combination of both during the course of the task (Scullin et al., 2013). When the appropriate moment to act is signalled by a PM cue it is a so-called event-based PM task (e.g., the exercise symbol on a page in the textbook). In case of a visual PM cue, the eyes need to fixate the PM cue in order to recognise it as distinct stimulus (in a pool of text and other symbols) and to identify the object (i.e. this is the exercise symbol). Graf (2005) emphasised the necessity of focally attending the PM cue in order to evoke processes that lead to a PM success. Following the noticing as distinct stimulus (see also Marsh et al., 2002), the object needs to be recognised as PM cue, i.e. as something special being important to the intended plan (Brandimonte and Passolunghi, 1994; Graf, 2005; Knight, 1998; Marsh et al., 2002). The PM cue and its context need to be verified (Marsh et al., 2003). The intention needs to be recalled or brought into consciousness (Graf, 2005; Knight, 1998). Furthermore, PM task and ongoing task need to be coordinated (Marsh et al., 2003), in the sense that the ongoing task is inhibited to flexibly switch to the PM task (Kliegel et al., 2002) allowing to initiate and to execute the intended action (Ellis, 1996).

According to the argumentation above, it is possible to distinguish between PM processes before attending the PM cue and those that begin once the PM cue receives focal attention. The current study used eye movements, more precisely fixations, to define the time intervals and analyse processes therein. Eye movements seem an ideal measure for PM processes, since they are influenced by bottom-up (e.g. stimulus content) as well as top-down aspects, such as task requirements (Land and Hayhoe, 2001), processing demands (e.g. Graupner et al., 2011; Yarbus, 1967) and adopted strategies (Green et al., 2007). Eye movement measures of fixation duration and saccadic amplitude reach adult-like values in the age group of 6–8 year olds (Helo et al., 2014) whereby no baseline differences are expected (we additionally use a visual search task as control). Gaze behaviour can be examined in terms of spatial distribution (e.g. fixation location) and temporal aspects (e.g. fixation duration; Mills et al., 2011). Most of the time the direction of visual attention corresponds to the location of the fixation (Henderson, 2007) meaning that the area in a picture which is fixated matches the processed content. Additionally, successful problem solving is accompanied by longer fixations on relevant problem features (Knoblich et al., 2001). Eye movements therefore provide spatial and temporal insights into the processing of information with high temporal resolution and precise spatial accuracy. For instance, Pannasch and Velichkovsky (2009) showed that reactions towards stimuli change within the same fixation, i.e. at the scale of milliseconds. Moreover, eye movements can be understood as a continuous measure, allowing recognition of changes in resource allocation towards PM task and ongoing task even within a single trial (see Hartwig et al., 2013). Hence, the analysis of eye movements can go beyond trial by trial comparisons and allows analysing processes before and after attending the PM cue.

To make such analyses possible, we need to define particular time intervals for trials containing a PM cue. The first time interval from the stimulus onset until the start of the first fixation (focal attention) on the PM cue will be labelled Time to First Fixation. The second time interval starts with the first fixation on the PM cue and is terminated with the fulfilment of the intended action in that trial. In case of a miss trial (i.e. a non-successful PM task) this interval would last until the end of the trial. Henceforth this time interval will be labelled as Time to Reaction. Trials with fulfilled intention contain a third time interval starting after fulfilment and lasting until trial end. Henceforth this time interval will be labelled as Time following Reaction. For a schematic diagram of the time intervals in a PM cue hit trial see Fig. 1. For trials without PM cue we defined equivalent time intervals using a data-based cut-off for the first fixation on the PM cue, see Method section for details (2.5.2 Time intervals, p. 12).

As outlined above, Time to First Fixation might include varying processes which are described by different theories. According to the Preparatory Attention and Memory Processes theory, PM success is impossible without preparatory attentional processes (Smith, 2003; Smith and Bayen, 2004). Thus, non-automatic monitoring for the PM cue is a prerequisite for successful PM performance. The current study focuses on external monitoring, i.e. monitoring the environment to detect the PM cue. In our example this would mean a scanning of each new textbook page for the exercise symbol. This monitoring has been described in event-based PM tasks as “the voluntarily deployment of attentional resources to search for the target cue” (p. 107, Ballhausen et al., 2019). Such a search occurs before the detection of a PM cue, i.e. in trials with and without PM cue (Smith, 2003; Smith and Bayen, 2004).

Following the Multiprocess Framework (McDaniel and Einstein, 2000)—in contrast to the Preparatory Attention and Memory Processes theory—PM success can be achieved by relying on spontaneous processes to notice the PM cue e.g. when it is difficult to apply the cognitive resources for monitoring (e.g. an absorbing ongoing task). Therefore, PM can be successful without particularly searching for the PM cue but focusing on the ongoing task and spontaneously detecting the PM cue.

Finally, Time to First Fixation might include monitoring as well as reliance on spontaneous processes, dynamically varying during the course of the task. For instance, since monitoring over a longer period of time is exhausting, this strategic activity might decrease during the course of the task (Einstein et al., 2005). In their Dynamic Multiprocess Framework Scullin et al. (2013) proposed that the monitoring activity might rebound following a spontaneous PM retrieval as this event signals that the current context might contain PM cues (e.g. after having spontaneously noticed an exercise symbol and performing the intended action, one starts to strategically monitor the following pages to not miss the next one). Also allowing for temporal dynamics in monitoring behaviour is a view proposing an active attention allocation to PM task and ongoing task during task encoding based on metacognitive beliefs about task demands (Hicks et al., 2005; Lourenço et al., 2015; Marsh et al., 2005). Allocating more attention to the PM task when believing the PM task is difficult results in greater ongoing task interference, i.e. a higher ongoing task cost as indicated by longer reaction times (Hicks et al., 2005). However, the attention allocation view also proposes that attention allocation is flexible, allowing the attention allocated to PM task and ongoing task to be adjusted during the course of the ongoing task. For instance, Loft et al. (2008) showed that ongoing task interference decreased when no PM cue was shown over a long period of time (the ongoing task interference was, however, never entirely eliminated). Moreover, the attention dedicated to the PM task can be adjusted in a matter of trials: Lourenço et al. (2015) found a larger amount of attention allocated to the PM task following a PM cue hit. Such adjustments in attention allocation also depend on whether the actual ongoing task load experienced during task execution matches the assumed ongoing task load during intention encoding (see Loft et al., 2008). I.e. if the information in the textbook is much harder to understand than anticipated before reading, children might direct more and more attention to the text, leaving less to notice the exercise symbol. Since attentional monitoring, reliance on spontaneous noticing and changes in attention allocation are key factors in PM research and the ability to monitor has been shown to develop during childhood, the present research will further examine these processes.

Allocating focal attention to the PM cue—either due to monitoring or spontaneous noticing—does not guarantee PM success, because subsequent processes need to be successful as well. These subsequent processes include recognising the fixated cue as PM cue (e.g., in the exercise example it is necessary to notice that the exercise symbol signifies the appropriate moment to answer the study questions). Furthermore, the intended action needs to be recalled (e.g., not only remembering that something needs to be done but also recalling what is required). Likewise, PM task and ongoing task need to be coordinated, e.g., inhibiting the ongoing task in order to initiate and perform the intended action (e.g., although children might experience the information in the textbook as intriguing they still need to inhibit reading and switch to the exercise task).

With respect to the intention recall several processes have been suggested. Intention recall can be spontaneous, i.e. based on the reflexive associative memory system (Moscovitch, 1994), where the PM cue interacts with a memory trace and the associated information is delivered to consciousness “rapidly, obligatorily and with few cognitive resources” (p. 129, McDaniel and Einstein, 2000). Intention recall can also be a resource intensive process: the notice + search view (Einstein and McDaniel, 1996) proposes intention recall to be a directed and controlled memory search. All these processes take place during Time to Reaction, which starts with the first fixation on the PM cue and ends with the button press. While noticing the PM cue might be fast (see e.g., Goschke and Kuhl, 1993), all other processes (including possible spontaneous intention retrieval) require time (Marsh et al., 2002).

The differentiation between Time to First Fixation and Time to Reaction is not trivial but can be achieved by relying on eye movement parameters. However, a further differentiation of processes within Time to Reaction only on the basis of eye movement parameters is difficult, but can be achieved by relating these parameters to executive function measures (i.e. inhibition, flexibility and working memory). We expect that these measures are in direct connection with intention recall and PM task and ongoing task coordination and therefore might also explain possible differences between children of different age groups.

PM development seems to follow an inverted U-shape (Kliegel et al., 2008). Children with 2 years of age have been shown to be able to successfully remember a PM task. PM performance has been found to increase during childhood and adolescence until young adulthood (e.g., Kerns, 2000; Mackinlay et al., 2009; Passolunghi et al., 1995), but see Nigro et al. (2002) for contrary findings. The increasing PM performance during childhood has been attributed to increasingly efficient strategic monitoring (Ceci et al., 1988; Kerns, 2000) but might also be due to better intention recall abilities (Kliegel and Jäger, 2007; Smith et al., 2010; Zöllig et al., 2007).

Rapidly developing executive functions have been called the reason for increasing PM performance during childhood (for an overview see Mahy et al., 2014a). Executive functions that have been shown to influence PM performance during childhood are, among others, inhibition (e.g., Kvavilashvili et al., 2001; Mahy et al., 2014b; Wang et al., 2008), working memory (e.g., Kliegel et al., 2013; Mahy et al., 2014b) and the ability to flexibly switch between tasks (Kerns, 2000). It is assumed that the development of inhibition, working memory and flexibility begins in early childhood and continues into adolescence or even young adulthood (see Mahy et al., 2014a for a comprehensive overview). Let us illustrate the relevance of the three executive functions for PM performance and identify them in the described time intervals: Inhibition is necessary to suppress the ongoing task and to allow switching from ongoing task to PM task. This switching process also requires flexibility to a certain extent. Working memory is essential to monitor for the PM cue but also to recall the intention from long-term memory. According to Braver's (2012) Dual Mechanisms of Control Framework working memory capacity could contribute to the use of proactive control, i.e. “bias[ing] attention, perception and action systems in a goal-driven manner” (p. 2). Therefore, higher working memory capacity could result in a shorter duration of Time to First Fixation as children utilise monitoring behaviour. It could also result in shorter Time to Reaction due to a faster retrieval of the intended action if the intended action is maintained on a higher activation level. Therefore, inhibition and flexibility might influence the duration of as well as eye movements during Time to Reaction, whereas working memory should have an effect on Time to First Fixation as well as Time to Reaction. However, Mahy et al. (2014a) also discuss the necessity of inhibition in relation to the ongoing task. The level of significance of these executive functions for successful PM performance also depends on task related factors such as the perceived difficulty of PM task and ongoing task, ongoing task absorption, association of PM cue and intention, the timing of PM task and ongoing task, etc.

Previous research mainly focused on PM processes in relation to performance data. For instance, longer reaction times in PM task blocks (compared to ongoing task only blocks) have been interpreted in terms of monitoring behaviour (e.g., Cohen et al., 2008; Einstein et al., 2005; Hicks et al., 2005; Smith, 2003). Similarly, the development of PM in childhood has been examined based on performance data. Smith et al. (2010) referred obtained ongoing task costs (i.e. longer reaction times) in 7 and 10 year old children to monitoring behaviour but these ongoing task costs have also been explained in terms of retrieval processes (Einstein and McDaniel, 2010; Smith, 2010). Ongoing task costs can also occur following the noticing of the PM cue. Leigh and Marcovitch (2014) mentioned that “It is worthwhile to speculate what the cost on the ongoing task may represent in young children” (p. 33) and went on to speculate about monitoring and holding the PM instruction in mind. And Loft et al. (2008) point out as well “it must be acknowledged that it is difficult to pinpoint the precise nature and function of the cognitive mechanisms that give rise to task interference” (p. 147). Due to these difficulties in interpretation, other, continuous measures have been asked for (Harrison and Einstein, 2010).

The current study aims to analyse parameters that are closer to the processes of interest by measuring eye movements. We therefore presented movie stills of Shaun the Sheep to the participating children with three different tasks. First, in an ongoing task only block, each still presentation was followed by a snippet. Children had to decide whether this snippet belonged to the previous still or not. The eye movements during this ongoing task only block resemble scene inspection as only the still was to explore. Second, in the PM task block the same ongoing task was used. In addition to the ongoing task, children's PM task consisted of pressing a specified button whenever seeing a bobble cap. Third, in a visual search task block children had to press the same button as in the PM task block whenever seeing a bobble cap. This design allows for the following comparisons: Differences in eye movements between ongoing task only and PM task block can be attributed to the additional PM task. If the PM processes are entirely spontaneous (i.e. no monitoring, spontaneous retrieval of the intended action, etc.), eye movements in ongoing task only and the PM task block should be very similar. In contrast, if children monitor for the PM cue, i.e. strategically search for the PM cue in the PM task + ongoing task, eye movements during Time to First Fixation should be similar to those during the visual search task. Furthermore, during the PM task block and the visual search task, the PM cue and visual search target, respectively (i.e. the bobble cap) needs focal attention in order to be identified as distinct stimulus and to initiate the encoded action (i.e. the button press). Differences between the PM task block and the visual search task during Time to Reaction therefore indicate PM processes of recognising the PM cue's importance, PM cue and context verification, recalling the intention from long term storage and PM task and ongoing task coordination. Taken together, the current study is designed to answer the following questions: 1) Do children of both age groups monitor for the PM cue? 2) Are there changes in monitoring behaviour over time in both age groups? 3) How do PM processes change within PM cue trials and are there age differences in these changes? The following paragraphs give an overview of our hypotheses. For a detailed outline of all three research questions, the dependent variables analysed to answer these research questions and our operationalised hypotheses see Table 1 in the Method section on p. 14. For a graphic that includes the time intervals and the eye tracking data in these time intervals please refer to Fig. 7 on p. 25.

We test whether both age groups monitor during Time to First Fixation for the PM cue during the PM task block. In a first analysis we will therefore contrast trial types and expect the following results. During the ongoing task only we expect eye movements that are characteristic for scene inspection, whereas in the visual search task children will show eye movements specialised to find visual search targets as fast as possible. More precisely, longer fixation durations during Time to First Fixation are predicted for ongoing task only than for the visual search task (see Hartwig et al., 2013 for a differentiation between scene inspection and visual search based on fixation durations). If children monitor for the PM cue (regardless of the absence or presence of the PM cue) this will be characterised by short fixation durations during Time to First Fixation comparable to the visual search task. If children do not monitor, fixation durations during Time to First Fixation will be comparable to the ongoing task only block in PM task + ongoing task for PM cue hit trials and trials without PM cue (henceforth called PM story trials). We expect older children to sustain monitoring behaviour across the whole task block while younger children's monitoring should decrease over time, intermittently increasing after a PM cue hit (but not after a PM cue miss). To test this dynamic in monitoring behaviour, we divide each task block into three segments with an equal number of trials. To examine monitoring dynamics independent of PM cue reactions (i.e. PM cue hit/miss), we select only the PM story trials from the PM task block and compare fixation durations in Time to First Fixation between the first, second and third segment. We suppose a positive correlation between monitoring behaviour and the scores in the working memory test. Furthermore, we will analyse the influence of PM cue reaction (hit or miss) on fixation durations in Time to First Fixation. Therefore, fixation durations in PM cue hit and miss trials will be contrasted with the two trials before and following the PM cue trial.

Since PM is assumed to be a multiple component process, we expect PM processes to change in the course of single PM cue trials: we expect monitoring to stop with the first fixation on the PM cue, and occurrence of processes of object identification, PM cue and context verification, intention recall, ongoing task inhibition and flexible switching during Time to Reaction. These process changes across time intervals should be characterised by longer fixation durations during Time to Reaction compared to Time to First Fixation in PM cue hit but not PM cue miss trials. Such an increase should exceed the often reported prolongation of fixations during scene inspection (see e.g. Pannasch et al., 2008) as controlled by the fixation duration during Time following Reaction.

We predict a longer duration of Time to Reaction as well as longer fixation durations during Time to Reaction for PM cue hit compared to visual search target hit trials. The assumed prolongation represents the additional PM processes (e.g. intention recall from long-term storage, PM task and ongoing task coordination). Also, the difficulty of underlying processes can be measured by the time spent fixating an object (Henderson, 1992; Rayner, 1978).

However, the prolongation of Time to Reaction should be less pronounced in older compared to younger children due to better inhibition, flexibility and working memory abilities. Should only some or none of the collected executive function scores explain Time to Reaction variance, our results might inform about the executive function dependence of PM processes during Time to Reaction in the tested age groups and paradigm.

Section snippets

Participants

We recruited 121 children from university staff, children's university, community daycare centres and schools. Of those children 10 had to be excluded due to experimenter error; six did not meet the visual acuity criterion of > 1.00 (equals 20/20 vision) as tested with the Freiburg Visual Acuity Test (FrACT; Bach, 2007); one did not understand the PM task; one had Aspergers Syndrome; one had false alarms in more than 1/3 of trials in the PM task. Nine were excluded due to their fidgetiness,

Executive function measures

In the flexibility task as well as the Go/Nogo task of the KiTAP older children made less errors than younger ones, χ2(1) = 12.72, p < .001, rate ratio = 1.949 and χ2(1) = 7.34, p = .007, rate ratio = 1.888, respectively. Furthermore, older children reached significantly higher overall scores in the digit span backwards task than younger ones, χ2(1) = 14.24, p < .001, rate ratio = 1.202 (see Table 2).

Correlating PM performance with these executive function measures showed a significant result

Discussion

The current experiment analysed and compared PM processes in children of two different age groups. In particular, we examined dynamics in PM processes over the course of a PM task block as well as within PM cue trials. Therefore, we measured eye movements while children completed the following task blocks: an ongoing task only block, a PM task block and a visual search task block, consecutively. We additionally captured inhibition, flexibility and working memory and included the performance

Funds

Funding for this study was provided by the Deutsche Forschungsgemeinschaft (German Research Foundation; DFG grants SFB 940/2, PA 1232/8-1).

Declaration of competing interest

None.

Acknowledgements

We thank Stefan Vogt, Johanna Fritz, Jonathan Wehnert, Julia Stietz, Mathias Kammerer, Lena Riese, and Max Schulz, for help with data collection.

References (102)

  • A.W. Inhoff et al.

    Definition and computation of oculomotor measures in the study of cognitive processes

  • L. Itti et al.

    A saliency-based search mechanism for overt and covert shifts of visual attention

    Vision Research

    (2000)
  • M. Kliegel et al.

    The effects of age and cue-action reminders on event-based prospective memory performance in preschoolers

    Cogn. Dev.

    (2007)
  • M. Kliegel et al.

    The role of noticing in prospective memory forgetting

    Int. J. Psychophysiol.

    (2007)
  • M. Kliegel et al.

    The development of prospective memory in young schoolchildren: the impact of ongoing task absorption, cue salience, and cue centrality

    J. Exp. Child Psychol.

    (2013)
  • M.F. Land et al.

    In what ways do eye movements contribute to everyday activities?

    Vis. Res.

    (2001)
  • J. Leigh et al.

    The cognitive cost of event-based prospective memory in children

    J. Exp. Child Psychol.

    (2014)
  • R.J. Mackinlay et al.

    Predictors of time-based prospective memory in children

    J. Exp. Child Psychol.

    (2009)
  • C.E.V. Mahy et al.

    Executive functioning and prospective memory in young children

    Cogn. Dev.

    (2011)
  • C.E.V. Mahy et al.

    The development of prospective memory in children: an executive framework

    Dev. Rev.

    (2014)
  • C.E.V. Mahy et al.

    The impact of age, ongoing task difficulty, and cue salience on preschoolers’ prospective memory performance: the role of executive function

    J. Exp. Child Psychol.

    (2014)
  • T. Mäntylä et al.

    Time monitoring and executive functioning in children and adults

    J. Exp. Child Psychol.

    (2007)
  • M.K. Scullin et al.

    The dynamic multiprocess framework: evidence from prospective memory with contextual variability

    Cogn. Psychol.

    (2013)
  • P.M. Van Diepen et al.

    Chronometry of foveal information extraction during scene perception

  • J. Allaire et al.

    rmarkdown: dynamic documents for R

  • M. Altgassen et al.

    Future thinking instructions improve prospective memory performance in adolescents

    Child Neuropsychol

    (2017)
  • M. Bach

    The Freiburg visual acuity test-variability unchanged by post-hoc re-analysis

    Graefes Arch. Clin. Exp. Ophthalmol.

    (2007)
  • A. Baddeley

    Exploring the central executive

    The Quarterly Journal of Experimental Psychology Section A

    (1996)
  • K. Barton

    MuMIn: multi-model inference

  • D.M. Bates

    lme4: Mixed-effects Modeling with R: Berlin

    (2010)
  • I. Biederman

    On the semantics of a glance at a scene

  • W. Bleidorn et al.

    Ein Big Five-Inventar für Kinder und Jugendliche

    Diagnostica

    (2009)
  • M.A. Brandimonte et al.

    The effect of cue-familiarity, cue-distinctiveness, and retention interval on prospective remembering

    The Quarterly Journal of Experimental Psychology Section A

    (1994)
  • S.J. Ceci et al.

    Don’t forget to take the cupcakes out of the oven’: prospective memory, strategic time-monitoring, and context

    Child Dev.

    (1985)
  • S.J. Ceci et al.

    Prospective remembering, temporal calibration, and context

  • A.-L. Cohen et al.

    Number of cues influences the cost of remembering to remember

    Mem. Cogn.

    (2008)
  • G.O. Einstein et al.

    Retrieval processes in prospective memory: theoretical approaches and some new empirical findings

  • G.O. Einstein et al.

    Prospective memory and what costs do not reveal about retrieval processes: a commentary on Smith, Hunt, McVay, and McConnell (2007)

    J. Exp. Psychol. Learn. Mem. Cogn.

    (2010)
  • G.O. Einstein et al.

    Multiple processes in prospective memory retrieval: factors determining monitoring versus spontaneous retrieval

    J. Exp. Psychol. Gen.

    (2005)
  • J. Ellis et al.

    Prospective memory or the realization of delayed intentions: A conceptual framework for research

  • J. Ellis et al.

    Prospective memory in 2000: past, present, and future directions

    Appl. Cogn. Psychol.

    (2000)
  • N. Galley et al.

    Fixation Durations–Why are they So Highly Variable? Advances in Visual Perception Research

    (2015)
  • A. Gelman et al.

    Data Analysis Using Regression and Hierarchical/Multilevel Models

    (2007)
  • T. Goschke et al.

    Representation of intentions: persisting activation in memory

    J. Exp. Psychol. Learn. Mem. Cogn.

    (1993)
  • P. Graf

    Prospective memory retrieval revisited

  • U. Halekoh et al.

    A Kenward-Roger approximation and parametric bootstrap methods for tests in linear mixed models – the {R} package {pbkrtest}

    J. Stat. Softw.

    (2014)
  • T.L. Harrison et al.

    Prospective memory: are preparatory attentional processes necessary for a single focal cue?

    Mem. Cogn.

    (2010)
  • J.M. Henderson

    Object identification in context: the visual processing of natural scenes

    Canadian Journal of Psychology/Revue canadienne de psychologie

    (1992)
  • J.M. Henderson

    Regarding scenes

    Curr. Dir. Psychol. Sci.

    (2007)
  • A. Hollingworth et al.

    Does consistent scene context facilitate object perception?

    J. Exp. Psychol. Gen.

    (1998)
  • Cited by (5)

    • Prospective memory performance and its improvement in individuals with high schizotypal traits: Evidence from eye-tracking studies

      2022, Clinical Neurophysiology
      Citation Excerpt :

      The stage after the occurrence of PM cues includes complex psychological processes, such as PM cue detection, intention retrieval and intention execution. Based on Hartwig et al. (2021), we made a fine division of the time from the occurrence of PM cues to the PM response and extracted four eye movement indices (i.e., time to first fixation, first fixation duration, total fixation duration and time from first fixation to response). Analysis of PM trial data revealed that implementation intention did not impact eye-movement indices between individuals with HSTs with implementation intention and individuals with HSTs with typical instruction.

    • LSTM-based Eye-Movement Trajectory Analysis for Reading Behavior Classification

      2022, Proceedings of SPIE - The International Society for Optical Engineering
    View full text