Bidirectional interference between timing and concurrent memory processing in children

Journal of Experimental Child Psychology 106 (2010) 145–162

Contents lists available at ScienceDirect

Journal of Experimental Child Psychology journal homepage: www.elsevier.com/locate/jecp

Bidirectional interference between timing and concurrent memory processing in children Anne-Claire Rattat Toulouse University–CUFR Jean-Francois Champollion, UTM, Octogone–ECCD, EA 4156, 81012 Albi Cedex 9, France

a r t i c l e

i n f o

Article history: Received 17 December 2009 Revised 2 February 2010 Available online 7 March 2010 Keywords: Timing Time Working memory Children Dual-task paradigm Interference effects

a b s t r a c t This study investigated the nature of resources involved in duration processing in 5- and 8-year-olds. The children were asked to reproduce the duration of a visual or auditory stimulus. They performed this task either alone or concurrently with an executive task (Experiment 1) or with a digit or visuospatial memory task (Experiment 2). The results showed that duration reproduction was systematically shorter in the dual-task condition than in the single-task one. Furthermore, timing an auditory stimulus decreased the proportion of accurate responses in the executive and digit memory tasks but not in the visuospatial memory task, whereas timing a visual stimulus decreased the proportion of accurate responses in the executive and visuospatial memory tasks but not in the digit memory task, at least to a lesser extent in the older children. This pattern of interference suggests that duration reproduction in children requires both the central executive and the slave memory system associated with the modality of the temporal stimulus. Ó 2010 Elsevier Inc. All rights reserved.

Introduction The underlying processes of time estimation have always intrigued psychologists (James, 1890), certainly because time is one of the dimensions that govern much of human behavior. Most timing researchers assume that humans possess an internal clock-like system that they can use to measure time (for a review, see Wearden, 2005). Moreover, recent studies have suggested that this internal clock is functional at a very early age (Brannon, Roussel, Meck, & Woldorff, 2004; Droit-Volet, 2003a; Droit-Volet, Clément, & Fayol, 2003; Droit-Volet & Wearden, 2001; see also Droit-Volet,

E-mail address: [email protected] 0022-0965/$ - see front matter Ó 2010 Elsevier Inc. All rights reserved. doi:10.1016/j.jecp.2010.02.001

146

A.-C. Rattat / Journal of Experimental Child Psychology 106 (2010) 145–162

2003b, for a review). Although the exact nature of this internal mechanism is still not fully understood, especially with regard to the brain structures involved (Lewis & Miall, 2003; Meck & Malapani, 2004), there is a general agreement that it involves an accumulator system. More specifically, in classic internal clock models (e.g., Gibbon, Church, & Meck, 1984), a pacemaker continuously emits pulses that are accumulated into an accumulator through an attention-controlled switch that opens and closes at the beginning and end of the estimated duration. The number of pulses accumulated during the event being timed is the internal representation of the event duration; the more pulses that are accumulated, the longer the perceived duration. However, time judgment also relies on memory and decisional processes. Indeed, the internal representation of the event duration is then stored in shortterm working memory and can then be transferred into long-term reference memory, which stores important times (e.g., standard durations important for the current task). Final temporal judgment results from a comparison between the just-presented duration and the durations stored in reference memory. According to this theoretical approach, duration is processed just like any other type of information. As such, its processing requires cognitive resources. This idea has been extensively supported by dual-task studies showing that parallel processing of a temporal task and a nontemporal task impairs duration estimation, leading participants to systematically underestimate durations (e.g., Brown & Merchant, 2007; Casini & Macar, 1997; Champagne & Fortin, 2008; Kladopoulos, Hemmes, & Brown, 2004; Predebon, 1996; Sawyer, 2003). This subjective shortening of perceived durations under dualtask conditions is widely taken as evidence that temporal processing requires working memory resources that are also needed by the concurrent nontemporal task. However, although it is well established that temporal information processing draws on working memory resources, much remains to be uncovered about the nature of these resources. Identifying the resources involved in time estimation is an important issue within the context of a multiple-resource memory model such as Baddeley and Hitch’s (1974) original working memory model. According to this model, working memory comprises three components: the central executive, phonological loop, and visuospatial sketchpad. The latter two components are slave systems specializing in processing and handling limited amounts of verbal or visuospatial information-typical images. In contrast, the central executive is an attentional controller mechanism responsible for a range of high-level functions, including the sharing of attentional resources between the two slave systems and the coordination and scheduling of processes in dual-task situations. In classic internal clock models, time refers to a continuous flow of information that can be tracked only by dint of constant attentional effort. Thus, we can logically expect the central executive component of working memory to play a role in temporal information processing. Recent studies have yielded empirical data that are consistent with this assumption (Brown, 1997, 2006; Brown & Frieh, 2000; Fortin, Champagne, & Poirier, 2007; Rammsayer & Ulrich, 2005). Brown (2006), for example, investigated bidirectional interference by pairing a timing task with an executive-level task. In his study, the timing task consisted in generating a series of 5-s temporal productions by pressing a button, whereas the central executive task consisted in verbally producing a continuous series of random numbers. In the dual-task condition, the adults’ temporal productions were longer and more variable, and the randomization in the nontemporal task was poorer, compared with the single-task condition. Brown also suggested that adults’ time estimation relies on executive-level resources rather than on resources dedicated to other specialized task demands such as visual and phonological processing. However, recent data from adults raise questions about whether the processing of temporal information necessarily requires attentional control monitoring by the central executive. Franssen, Vandierendonck, and Van Hiel (2006) found that a task relying on the phonological loop without attracting attentional resources also affected time estimation. In their study, adults needed to reproduce or verbally estimate short tone durations (<4 s) in different phonological load-level conditions, namely no load, performing articulatory suppression, and listening to irrelevant speech, tones, or music. The results showed that articulatory suppression alone affected timing performance (i.e., the durations were systematically underestimated), suggesting that adults’ time estimation is mediated by phonological working memory and the involvement of an active verbal rehearsal process. Nevertheless, the authors mentioned that it was still a matter of debate whether the concurrent active information processing in memory that leads to impaired time estimation also requires attentional

A.-C. Rattat / Journal of Experimental Child Psychology 106 (2010) 145–162

147

control by the central executive. Moreover, in their study, all of the stimuli to be timed were tones, thereby making it impossible to establish whether presenting the temporal stimuli in the visual modality might yield different results. At first glance, it would appear to be logical for time estimation to be associated primarily with the central executive component of working memory insofar as classic internal clock models state that prospective timekeeping requires continuous monitoring and updating of the passage of time. However, the work by Franssen and colleagues (2006) demonstrates the need for further research on the involvement of the other working memory subsystems in time estimation. Put differently, the question is whether duration processing relies solely on executive resources, insofar as duration is nonmodal information, or whether slave memory mechanisms are also involved in time processing to a greater or lesser extent, possibly depending on whether the stimulus to be timed was presented in the auditory or visual modality. Indeed, Franssen and colleagues (2006) and previously Wearden and Culpin (1995, 1998), provided experimental data suggesting that the processing of auditory durations relies on phonological resources. Wearden, Parry, and Stamp (2002) also put forward the possibility that stimuli that cannot be phonologically encoded may instead be stored in the visuospatial sketchpad subsystem of the working memory. However, to date, no clear evidence that the processing of visual durations involves the visuospatial sketchpad memory system has been found. The purpose of the current study was to examine the nature of the resources dedicated to auditory and visual timing further by using a developmental approach in children—something that had never been done before. Previous studies have been conducted mainly with adults, and little is known about the developmental aspects of young children’s short-term memory for durations (Droit-Volet, Wearden, & Delgado, 2007). Note that until now research related to memory development for time has focused mainly on different aspects of long-term memory in children, such as temporal order (e.g., Bauer & Shore, 1987) and temporal location (e.g., Tartas, 2001), and more recently has focused on recall of event duration (e.g., Rattat & Droit-Volet, 2005, 2007). Thus, it is important to specify that the developmental approach used in our study allowed us to make several clear predictions about the effects of concurrent memory processing on temporal performance (see below) and, consequently, to address the fundamental question of the extent to which different types of cognitive resources are involved in time processes. In the developmental literature on memory, there is considerable agreement that the amount of available cognitive resources increases from the preschool years to adolescence (for reviews, see Cowan, 1997, 2005; Gathercole, 1998, 2002; Towse, Hitch, & Horton, 2007). However, to our knowledge, the extent to which children’s limited memory capacity affects time estimation in dualtask conditions similar to those used in adult research has never been examined. Arlin (1986a, 1986b) and more recently Gautier and Droit-Volet (2002) are the only researchers to have examined young children’s time estimation in classic dual-task conditions. These authors asked 5- and 8-year-olds to reproduce a 6- or 12-s stimulus duration in either a single-task condition or a dual-task condition with a secondary task that consisted of naming a series of pictures presented at regular intervals. The results showed that, as in adults, adding a nontemporal task had no significant effect on the variability of time judgments, whereas it affected their mean by producing a shortening effect. This shortening effect in the dual-task condition was nevertheless smaller at 8 years of age than at 5 years of age due to the greater amount of resources available to the older children. However, although the age-related interference effect resulted from the general increase in available resources, there remained the question of why nontemporal performance in the dual-task condition was impaired by the temporal task (increased proportion of picture-naming errors) to the same extent in both age groups. Gautier and Droit-Volet speculated that, because tracking the passage of time is attentionally effortful, the 5-year-olds gave priority to the nontemporal task in the dual-task condition. A possible alternative explanation for this pattern of results is that the workload demands of temporal and nontemporal tasks are managed, at least in part, by resources from different resource pools. A common interpretation of asymmetric patterns of dual-task interference is that there is a partial overlap in the resources that support the two tasks (e.g., Brown, 1997, 2006). However, the nontemporal picture-naming task used in Gautier and Droit-Volet’s (2002) study mixed different types of information, namely verbal and visual information, thereby preventing any dissociation between the working memory components involved when a temporal task is performed.

148

A.-C. Rattat / Journal of Experimental Child Psychology 106 (2010) 145–162

In the light of these considerations, the current study focused on the involvement of the central executive (Experiment 1) and the two slave (Experiment 2) memory components in time reproduction by 5- and 8-year-olds. Thus, the children were asked to reproduce stimulus durations lasting 6 or 12 s—the two durations used by Gautier and Droit-Volet (2002)—presented in two sensory modalities (visual or auditory). This task was performed either alone or concurrently with a memory task. We examined whether a memory task would impair temporal performance and, conversely, whether a time reproduction task would impair nontemporal memory performance. In our first experiment, the concurrent memory task was the verbal version of the random two-choice reaction time (CRTR) (Szmalec, Vandierendonck, & Kemps, 2005) recently used by Imbo, Vandierendonck, and Vergauwe (2007), which interferes with executive functions without placing an excessive load on the verbal and visuospatial slave components of working memory. Thus, if executive-level resources are allocated to time reproduction, we would expect to observe bidirectional interference in a dual-task condition. In other words, adding a central executive task should result in a temporal underestimation of the stimuli being timed regardless of whether the stimulus is presented in the visual or auditory modality. In contrast, as in Gautier and Droit-Volet’s (2002) study, no change in the variability of time judgments was expected. Conversely, a concurrent timing task should also disrupt nontemporal performance by making responses more random. Moreover, because of their lower central executive capacity, these bidirectional interference effects should be greater in the 5-year-olds than in the 8-year-olds. In our second experiment, the involvement of the two slave memory systems in children’s time reproduction was investigated by using two concurrent memory tasks, namely a classic digit memory task that primarily taxed phonological resources and an adaptation of the Corsi’s block-tapping task that primarily taxed visuospatial resources. Regardless of memory task and duration modality, we would expect that children’s temporal reproduction would be shorter, but not more variable, in the dual-task condition than in the single-task one. More important, adding a temporal task should in return disrupt nontemporal performance for a working memory task, but only if the same memory system is used for two tasks (visual modality duration + visuospatial memory task or auditory modality duration + digit memory task). Moreover, as in our first experiment, these bidirectional interference effects should be more pronounced in the younger children than in the older children due to the quantitative improvement of phonological and visuospatial memory during childhood.

Experiment 1 Method Participants and design A total of 54 children took part in this experiment: 26.5-year-olds (11 girls and 15 boys, mean age = 5.48 years, SD = 0.38) and 28.8-year-olds (15 girls and 13 boys, mean age = 8.15 years, SD = 0.30). Children were recruited (with parental consent) from nursery and primary schools in Lyon and Toulouse, France. In each age group, children were randomly assigned to one of the two duration modality conditions, visual (n = 26) or auditory (n = 28), in which the temporal reproduction task was performed either under a single-task condition or concurrently with the executive-level task. The executive-level task was also performed under a single-task condition. This within-participants design was randomly counterbalanced.

Materials A Power Macintosh computer, placed approximately 50 cm from the child, controlled all of the experimental events and recorded data in the temporal task via the PsyScope program (Cohen, MacWhinney, Flatt, & Provost, 1993). The visual temporal stimulus was a blue circle, 15 cm in diameter, displayed against a black background in the center of the computer screen. The auditory temporal stimulus was a 500-Hz tone produced by the computer speakers. For the temporal task, responses were made by pressing the space bar of the computer keyboard twice: once for the beginning and once for the end. For the nontemporal task that primarily taxed central executive resources, namely the

A.-C. Rattat / Journal of Experimental Child Psychology 106 (2010) 145–162

149

CRT-R task, the stimuli were low (262-Hz) and high (524-Hz) tones produced by the speakers. The experimenter checked children’s responses online, deciding whether they were correct or incorrect. Procedure The experiment lasted 25 min on average. Each child was tested individually in a quiet room and performed the following three tasks in random order: (a) a single temporal task (i.e., temporal reproduction), (b) a single central executive task, and (c) the temporal task with the central executive task (dual-task condition). Before each task, the child was given three demonstration trials and three practice trials. Note that five 5-year-olds were excluded from the final sample because, despite the practice trials, they failed to produce a higher mean response duration for the longer target duration than for the shorter one. In each trial of the single temporal task, the children were presented with a 6- or 12-s temporal stimulus (blue circle or tone). They were then instructed to reproduce the duration of this stimulus by pressing the space bar once to mark the start and once when they judged that sufficient time had elapsed. All children underwent 20 trials, with 10 trials for each target duration (6 or 12 s), presented in random order. In addition, children were explicitly told not to count. Note that it is well known that children younger than 7 or 8 years do not spontaneously use a counting strategy to time continuous durations (Wilkening, Levin, & Druyan, 1987). In the single central executive task (i.e., CRT-R task), a series of low (262-Hz) and high (524-Hz) tones was played. Each tone lasted 200 ms, and the interval between two successive tones was either 900 or 1500 ms. After each tone, the children needed to say ‘‘haut” (high) when they heard a high tone or ‘‘bas” (low) when they heard a low tone (for a similar procedure, see Imbo et al., 2007). In the dual-task condition (i.e., temporal reproduction + central executive task), the children were asked to perform a central executive task during the presentation of the stimulus duration to be reproduced. As for the single temporal task condition, children were instructed to reproduce the duration of the temporal stimulus immediately after its presentation. The central executive task was similar to that described above. For the 6-s trials children heard a series of three tones, and for the 12-s trials they heard a series of six tones, in random order. Each child performed 10 trials in the dual-task condition, with 5 trials per target duration. A press on the space bar initiated each trial. Results Temporal reproduction task: Mean response duration and reproduction variability Children’s responses in the temporal reproduction task were converted into two usual measures for analysis, namely the mean response duration and the coefficient of variation (e.g., Droit-Volet & Rattat, 1999; Rattat & Droit-Volet, 2002; Ulbrich, Churan, Fink, & Wittmann, 2007). Fig. 1 shows the mean response durations for the three age groups (5-year-olds, 8-year-olds, and adults) as a function of target duration (6 s vs. 12 s) and task condition (single task vs. dual task) for the visual (upper panel) and auditory (lower panel) modalities used for the temporal stimulus. An examination of this figure suggests that mean response durations were shorter in the dual-task condition than in the single-task one for both duration modalities. Moreover, this decrease in mean response durations in the dual-task condition compared with the single-task one was greater for the 12-s stimulus than for the 6-s one. This was confirmed by the statistical analyses. Mean response duration Previous analyses revealed neither a significant main effect nor any interaction effect involving the button order and task order factors. Thus, these factors were not included in the statistical analyses. An analysis of variance (ANOVA) was conducted on mean response durations with two between-participants factors (age and modality) and two within-participants factors (task condition and target duration). The ANOVA revealed significant main effects of target duration, F(1, 50) = 396.29, p < .0001, g2p = .89, and task condition, F(1, 50) = 135.09, p < .0001, g2p = .73, as well as a significant interaction between the two, F(1, 50) = 43.03, p < .0001, g2p = .46. As expected, the latter indicates that the children reproduced longer durations in the single-task condition than in the dual-task one, but to a greater extent for the 12-s target duration than for the 6-s one, t(53) = 6.19, p < .0001. There was no

150

A.-C. Rattat / Journal of Experimental Child Psychology 106 (2010) 145–162

Fig. 1. Mean response durations reproduced by the 5- and 8-year-olds for the visual (upper panel) and auditory (lower panel) stimulus modalities as a function of target duration (6 or 12 s) and task condition (single task or dual task) in Experiment 1.

significant main effect of modality, F < 1. Although the main effect of age was not significant, F < 1, there was a significant Task Condition  Target Duration  Age interaction, F(1, 50) = 7.11, p = .01, g2p = .13. There was no other significant interaction effect. To examine more closely the significant three-way interaction among task condition, target duration, and age, we calculated for each child the difference between mean response durations reproduced in the single- and dual-task conditions for each target duration. This difference was greater

151

A.-C. Rattat / Journal of Experimental Child Psychology 106 (2010) 145–162

in the 5-year-olds than in the 8-year-olds for the 12-s target duration, t(52) = 2.03, p = .047, but not for the 6-s target duration, t < 1.

Variability To investigate temporal reproduction variability, we calculated the coefficient of variation, which corresponds to the ratio of the standard deviation to the mean response duration (Table 1). An ANOVA conducted on the coefficient of variation, with age and modality as between-participants factors and task condition and target duration as within-participants factors, revealed a significant main effect of age, F(1, 50) = 18.08, p < .0001, g2p = .27. Consistent with the results of developmental studies showing that reproduction variability decreased with age (e.g., Droit-Volet & Rattat, 1999; Gautier & Droit-Volet, 2002), the reproduction variability was higher in the 5-year-olds than in the 8-year-olds (.39 vs. .26). There was also a significant Age  Target Duration interaction effect, F(1, 50) = 12.77, p = .001, g2p = .20. Reproduction variability was greater in the 5-year-olds than in the 8-year-olds for both the shorter target duration (.35 vs. .28), t(52) = 2.11, p = .04, and the longer target duration (.42 vs. .24), t(37.49) = 2.83, p = .009 (because the Levene test for equality of variances indicated that the assumption of homogeneity was violated for the longer target duration, the degrees of freedom were adjusted to account for this violation during statistical analysis). It should also be noted that reproduction variability was greater for the 12-s target duration than for the 6-s one in the 5-year-olds (.42 vs. .35), t(25) = 2.40, p = .024. The reverse was observed for the 8-year-olds, with reproduction variability being lower for the 12-s target duration than for the 6-s one (.24 vs. .28), t(27) = 2.83, p = .009. There was no other significant main or interaction effect. Therefore, as expected, the dual-task condition shortened the mean response duration in the two age groups but had no significant effect on time reproduction variability.

Central executive task (CRT-R task): Proportion of accurate responses An ANOVA was carried out on the proportion of accurate responses (i.e., correct identifications of the low or high frequency of tones) with age and modality as between-participants factors and task condition (single-task vs. 6-s dual task vs. 12-s dual task) as a within-participants factor. As expected, the main effect of modality was not significant, F(1, 50) = 2.32, p = .134, and there was no significant interaction effect involving this factor. There was a significant main effect of age, F(1, 50) = 20.23, p < .0001, g2p = .29. As can be clearly seen in Fig. 2, the proportion of accurate responses was lower in the 5-year-olds than in the 8-year-olds (.66 vs. .80, p < .0001). The main effect of task condition, F(2, 100) = 17.49, p < .0001, g2p = .26, was also significant. A posteriori comparisons using the Bonferroni adjustment showed that the proportion of accurate responses was higher in the single-task condition than in both the 6- and 12-s dual-task conditions (.79 vs. .71, p = .001, and .78 vs. .69, p < .0001, respectively), whereas no difference emerged when the two dual-task conditions were compared (.71 vs. .69, p = .579). Nevertheless, planned comparisons revealed that the decrease in the proportion of correct responses between the single- and dual-task conditions was greater in the younger children than in the 8-year-olds for the longer target duration ( .12 vs. .06), t(52) = 2.05, p = .045, but not for the shorter one ( .10 vs. .05), t(52) = 1.35, p = .182.

Table 1 Coefficients of variation for the 5- and 8-year-olds in the single- and dual-task conditions as a function of target duration (6 or 12 s) and stimulus presentation modality (visual or auditory) in Experiment 1. Stimulus modality

Single task

Dual task

Visual

Auditory

Visual

Auditory

Duration 6s

Age group 5-year-olds 8-year-olds

.31 .31

.34 .24

.31 .33

.42 .24

12 s

5-year-olds 8-year-olds

.43 .27

.42 .23

.43 .29

.40 .18

152

A.-C. Rattat / Journal of Experimental Child Psychology 106 (2010) 145–162

Fig. 2. Mean proportions of accurate responses provided in the central executive task by the 5- and 8-year-olds for the visual (upper panel) and auditory (lower panel) stimulus modalities as a function of task condition (single task, 6-s dual task, or 12-s dual task) in Experiment 1.

Discussion The results of Experiment 1 clearly show bidirectional interference between timing and concurrent central executive processing in children. More precisely, adding a concurrent executive-level task had no effect on time reproduction variability, although it did affect time accuracy by producing a shortening effect, especially in the younger children and especially for the duration of 12 s. More specifically, this shortening effect due to the concurrent executive processing was greater in the 5year-olds than in the 8-year-olds for the 12-s target duration, whereas it was similar in the two age

A.-C. Rattat / Journal of Experimental Child Psychology 106 (2010) 145–162

153

groups for the 6-s target duration. The greater shortening effect in the younger children for the longest duration may have been a consequence of the high processing demands imposed by this particular condition. It is easy to understand why tracking the passage of time for a long duration might require more sustained attentional effort. Relatively few studies have investigated time estimation in children using durations longer than 10 s, chiefly because the latter tend to find timing tasks to be rather boring. However, several studies have suggested that young children have greater difficulty in maintaining a constant attentional effort during time encoding (for a review, see Droit-Volet, Delgado, & Rattat, 2006). Conversely, the timing of stimuli decreased the proportion of accurate responses in the concurrent executive task in both age groups, but to a greater extent in the 5-year-olds for the longest target duration. Thus, these findings confirm and extend to young children those found in adults by Brown (2006) showing that performing an executive-level task and a timing task simultaneously degraded performance in both tasks. Moreover, our data show that the bidirectional interference between concurrent temporal and executive processing in children is similar regardless of the sensory modality in which the temporal stimulus is presented (visual or auditory). In sum, Experiment 1 provides support for the idea that timing involves executive resources and extends it to children as young as 5 years. Even though time reproduction is supported by executive resources, as explained in the Introduction, we cannot exclude the possibility that other types of working memory resources may also contribute to time reproduction depending on the modality of the presentation of the temporal stimulus. The objective of the second experiment below was to address this question by examining bidirectional interference effects between visual and auditory duration reproduction and two concurrent memory tasks that primarily taxed either phonological resources (a digit memory task) or visuospatial resources (an adaptation of the Corsi’s block-tapping task). In line with literature presented in the Introduction (e.g., Franssen et al., 2006; Wearden & Culpin, 1995, 1998; Wearden et al., 2002), we hypothesized that adding an auditory temporal task should disrupt performance in the digit memory task, whereas adding a visual temporal task should disrupt performance in the visuospatial memory task. Moreover, the classical shortening effect should also be observed regardless of stimulus modality and memory task. In addition, both of these interference effects should be more pronounced in the 5year-olds than in the 8-year-olds due to the higher phonological memory capacities (e.g., Hulme, Thomson, Muir, & Lawrence, 1984) and visuospatial memory capacities (e.g., Riggs, McTaggart, Simpson, & Freeman, 2006) in the older children. Experiment 2 Method Participants and design A total of 54 new children participated in this experiment: 27 5-year-olds (17 girls and 10 boys, mean age = 5.37 years, SD = 0.31) and 27 8-year-olds (17 girls and 10 boys, mean age = 8.49 years, SD = 0.29). They were recruited (with parental consent) from nursery and primary schools in Clermont-Ferrand and Toulouse, France. As in Experiment 1, in each age group, the children were randomly assigned to one of the two duration modality conditions: visual (n = 30) or auditory (n = 24). In each modality condition, the temporal reproduction task was performed either under a single-task condition or concurrently with the digit or visuospatial memory task. Thus, there were four dual tasks, namely the visual timing/visuospatial memory task, visual timing/digit memory task, auditory timing/ visuospatial memory task, and auditory timing/digit memory task. The digit and visuospatial memory tasks were also performed under a single-task condition. This within-participants design was randomly counterbalanced. Materials The materials used for the temporal task in the current experiment were the same as those used in Experiment 1. For the visuospatial memory task, a series of three small squares, which could be filled (black) or unfilled (white), was presented in the center of the computer screen, as explained below.

154

A.-C. Rattat / Journal of Experimental Child Psychology 106 (2010) 145–162

Note that in the dual task using the visual temporal stimuli and the visuospatial interfering task, the small squares were presented inside the blue circle so that the stimuli for both tasks could be viewed simultaneously. For the digit memory task, the computer speakers emitted a sequence of three digits randomly chosen between 1 and 8. For the visuospatial memory task as well as for the digit memory task, the number of presented items in one sequence (i.e., three items) corresponds to the median memory span from a group of 5-year-olds who were previously tested on the digit short-term memory span test. In this test, the experimenter enumerated a pseudorandom sequence of digits at the rate of approximately 1 per second. The children were also asked to repeat this sequence aloud. The difficulty level was manipulated

Fig. 3. Mean response durations reproduced by the 5- and 8-year-olds for the visual (upper panel) and auditory (lower panel) stimulus modalities as a function of target duration (6 or 12 s) and task condition (single task, digit dual task, or visuospatial dual task) in Experiment 2.

A.-C. Rattat / Journal of Experimental Child Psychology 106 (2010) 145–162

155

by increasing the number of digits included in the sequence. The digit score was calculated on the basis of the longest sequence that children repeated correctly. Procedure All children performed the five following tasks in a random order: three single tasks, namely (a) a single temporal task (i.e., temporal reproduction) that was strictly identical to that in the previous experiment, (b) a single digit memory task, and (c) a single visuospatial memory task, and two dual tasks, namely (d) the temporal task with the digit memory task and (e) the temporal task with the visuospatial memory task. As in Experiment 1, despite the three demonstration trials and the three practice trials preceding each task, several children (i.e., three 5-year-olds and one 8-year-old) were excluded from the final sample because they reproduced a mean duration that did not differ as a function of target duration and task condition. In the single digit memory task, children were given trials of 6 and 12 s. For the 6-s trials, they were presented with one digit sequence, and for the 12-s trials, they were presented with two digit sequences. For each digit sequence, children were presented with three digits randomly chosen between 1 and 8. Each digit was presented during 540 ms, and the interdigit interval was randomly chosen between 500 and 880 ms. The duration of a sequence was also 3 s. Then children needed to repeat this digit sequence aloud, and another trial (for the 6-s trials) or another sequence (for the 12-s trials) began 3 s after the previous sequence. A press on the space bar initiated each trial. Children performed 10 trials, 5 for the 6-s trials with one digit sequence and 5 for the 12-s trials with two digit sequences, in a random order. For the single visuospatial memory task, the procedure was similar to that for the digit memory task. However, in this task, three unfilled squares were presented on the computer screen and the children saw a sequence of three filled squares; that is, the squares temporarily changed color. Then children were asked to reproduce the filled squares sequence by touching the appropriate unfilled squares. In the two dual tasks (i.e., temporal reproduction + digit memory task and temporal reproduction + visuospatial memory task), children were asked to perform the interfering task during the presentation of the stimulus duration to be reproduced. As for the single temporal task condition, children were instructed to reproduce the duration of the temporal stimulus (6 or 12 s) immediately after its presentation. The interfering tasks (digit and visuospatial memory tasks) were similar to those described above. The children were given 10 trials for each dual task, 5 trials per target duration. Results Temporal reproduction task: Mean response duration and reproduction variability Mean response duration. As in Experiment 1, previous analyses revealed neither a significant main effect nor any interaction effect involving the button order and task order factors; therefore, these factors were not included in the statistical analyses. The ANOVA conducted on the mean response duration with two between-participants factors (age and modality) and two within-participants factors (task condition and target duration) revealed a main effect of age, F(1, 50) = 9.79, p = .003, g2p = .16, indicating that the 5-year-olds reproduced longer durations than the 8-year-olds (Fig. 3). The main effects of target duration, F(1, 50) = 335.38, p < .0001, g2p = .87, modality, F(1, 50) = 36.20, p < .0001, g2p = .42, and task condition, F(2, 100) = 59.55, p < .0001, g2p = .54, were also significant. Moreover, the Modality  Target Duration interaction, F(1, 50) = 6.47, p = .014, g2p = .12, the Modality  Task Condition interaction, F(2, 100) = 17.05, p < .0001, g2p = .25, the Task Condition  Target Duration interaction, F(2, 100) = 20.37, p < .0001, g2p = .29, and the Target Duration  Task Condition  Modality interaction, F(2, 100) = 6.38, p = .002, g2p = .11, were significant. There were no other significant interaction effects. To examine more closely the significant three-way interaction among modality, target duration, and task condition, we ran an ANOVA on the mean response duration for each stimulus duration modality separately. For the visual modality duration, there were significant main effects of target duration, F(1, 29) = 199.26, p < .0001, g2p = .87, and task condition, F(2, 58) = 63.01, p < .0001, g2p = .69, as well as an interaction effect between these two factors, F(2, 58) = 28.36, p < .0001, g2p = .49. The difference

156

A.-C. Rattat / Journal of Experimental Child Psychology 106 (2010) 145–162

between mean response durations reproduced in the single- and dual-task conditions differed from zero for both the visuospatial memory task (6 s, t(29) = 8.99, p < .0001, and 12 s, t(29) = 10.00, p < .0001) and the digit memory task (6 s, t(29) = 5.64, p < .0001, and 12 s, t(29) = 8.89, p < .0001). Put differently, regardless of the memory task, children reproduced longer durations in the single-task condition than in the dual-task one. This underestimation of durations in dual-task conditions compared with the single-task condition was nevertheless higher for the 12-s target duration than for the 6-s one: digit memory task, t(29) = 4.14, p = .001, and visuospatial memory task, t(29) = 5.82, p < .0001. Furthermore, for the 6-s target duration, there was no difference between the temporal underestimation in the dual task observed for the visuospatial and digit memory tasks, t < 1. It was only for the longest target duration, 12 s, that the difference between mean response durations reproduced in the single- and dual-task conditions was greater for the visuospatial memory dual-task condition than for the digit memory one, t(29) = 2.76, p = .01. As for the visual stimulus duration modality, the main effect of target duration was significant for the auditory modality, F(1, 23) = 142.78, p < .0001, g2p = .86, as was the main effect of task condition, F(2, 46) = 9.55, p < .0001, g2p = .29. The difference between mean response durations reproduced in the single and dual tasks differed from zero for both the visuospatial task, t(23) = 2.89, p = .008, and the digit memory task, t(23) = 5.81, p < .0001, with no difference being observed when the two dual tasks were compared, t < 1. Thus, for both target durations, the mean response duration was higher in the single task than in the two dual tasks. However, in contrast to the visual modality duration, there was no significant interaction effect between target duration and task condition for the auditory modality, F(2, 46) = 1.89, p = .163. Furthermore, the underestimation of duration between the single- and dual-task conditions was systematically higher for the visual stimulus modality than for the auditory one for both the visuospatial task, t(52) = 4.70, p < .0001, and the digit memory task, t(52) = 5.22, p < .0001.

Variability. As in Experiment 1, the ANOVA run on the coefficient of variation revealed significant main effects of age, F(1, 50) = 28.72, p < .0001, g2p = .37, and target duration, F(1, 50) = 4.86, p = .032, g2p = .37. The first result indicates that the reproduction variability was higher in the 5-year-olds than in the 8year-olds, and the second result indicates that it was higher for the 12-s target duration than for the 6s one (Table 2). The interaction between modality and task condition was also significant, F(2, 100) = 3.76, p = .027, g2p = .07. t Tests for paired samples revealed that for both stimulus modalities, the reproduction variability was similar in the single- and two dual-task conditions: visual, digit dual task, t < 1, and visuospatial dual task, t(29) = 1.27, p = .21; auditory, digit dual task, t(23) = 1.35, p = .19, and visuospatial dual task, t(23) = 1.73, p = .096. Moreover, for the single-task condition as well as for the digit dual-task condition, reproduction variability was similar for the visual and auditory stimulus duration, both t < 1. In contrast, reproduction variability was significantly higher for the visual stimulus duration than for the auditory one in the visuospatial dual-task condition (.32 vs. .24), t(49.74) = 3.05, p = .004 (because the Levene test for equality of variances indicated that the assumption of homogeneity was violated for the visuospatial dual-task condition, the degrees of freedom were adjusted to account for this violation during statistical analysis). Thus, as in the previous experiment, the dual-task conditions shortened the mean response duration in both the 5- and 8-year-olds, whereas they had no significant effect on time reproduction variability.

Table 2 Coefficients of variation for the 5- and 8-year-olds in the single-task, digit, and visuospatial dual-task conditions as a function of target duration (6 or 12 s) and stimulus presentation modality (visual or auditory) in Experiment 2. Stimulus modality

Digit dual task

Visuospatial dual task

Visual

Single task Auditory

Visual

Auditory

Visual

Auditory

Duration 6s

Age group 5-year-olds 8-year-olds

.36 .24

.33 .23

.34 .24

.33 .28

.38 .28

.28 .23

12 s

5-year-olds 8-year-olds

.36 .19

.29 .22

.33 .21

.33 .22

.34 .28

.24 .21

A.-C. Rattat / Journal of Experimental Child Psychology 106 (2010) 145–162

157

Digit and visuospatial memory tasks: Proportion of accurate responses For the digit and visuospatial memory tasks, a response was considered to be accurate when the child repeated the one- and two-item sequences without error for the 6- and 12-s target durations, respectively. The ANOVA carried out on the proportion of accurate responses with two between-participants factors of age and modality and two within-participants factors of memory task and task condition revealed a significant main effect of age, F(1, 50) = 91.04, p < .0001, g2p = .65, indicating a higher proportion of accurate responses in the older children than in the younger children (Fig. 4). The main effect of memory task, F(1, 50) = 63.28, p < .0001, g2p = .56, was also significant, suggesting that the proportion of accurate responses was higher for the digit memory task than for the visuospatial one. The main effect of task condition reached significance as well, F(2, 100) = 21.56, p < .0001, g2p = .30,

Fig. 4. Mean proportions of accurate responses provided in the digit and visuospatial memory tasks by the 5- and 8-year-olds for the visual (upper panel) and auditory (lower panel) stimulus modalities as a function of task condition (single task, 6-s dual task, or 12-s dual task) in Experiment 2.

158

A.-C. Rattat / Journal of Experimental Child Psychology 106 (2010) 145–162

whereas the main effect of modality did not, F(1, 50) = 1.29, p = .262. However, there was a significant four-way interaction among task condition, memory task, age, and modality, F(2, 100) = 4.93, p = .009, g2p = .09, as well as a significant Modality  Memory Task interaction, F(1, 50) = 10.20, p = .002, g2p = .17, a significant Task Condition  Memory Task interaction, F(2, 100) = 4.60, p = .012, g2p = .08, and a significant Task Condition  Memory Task  Modality interaction, F(2, 100) = 10.92, p < .0001, g2p = .18. No other significant interaction effects were observed. To examine the four-way interaction more closely, we performed an ANOVA on the proportion of accurate responses for each temporal stimulus modality separately. For the visual modality stimulus duration, the three main effects were significant: age, F(1, 28) = 26.62, p < .0001, g2p = .49, memory task, F(1, 28) = 53.92, p < .0001, g2p = .66, and task condition, F(2, 56) = 10.21, p < .0001, g2p = .27). The Task Condition  Memory Task interaction, F(2, 56) = 10.48, p < .0001, g2p = .27, and the Age  Task Condition  Memory Task interaction, F(2, 56) = 6.07, p = .004, g2p = .18, were also significant. The Task Condition  Age interaction, F < 1, and the Memory Task  Age interaction, F(1, 28) = 3.02, p = .09, did not reach statistical significance. Post hoc analyses revealed that at 5 years of age, for both the 6- and 12-s target durations, the proportion of accurate responses was higher in the single-task condition than in the dual-task one for the visuospatial memory task (6 s, t(13) = 3.54, p = .004, and 12 s, t(13) = 2.42, p = .031) but not for the digit memory task (6 s, t(13) = 1.35, p = .20, and 12 s, t < 1). In contrast, at 8 years of age, the proportion of accurate responses was higher in the single-task condition than in the dual-task one not only for the visuospatial memory task (6 s, t(15) = 2.33, p = .034, and 12 s, t(15) = 3.93, p = .001) but also for the digit memory task (6 s, t(15) = 2.07, p = .056, and 12 s, t(15) = 2.50, p = .024), albeit to a lesser extent. More precisely, the difference between the proportion of accurate responses obtained in the single- and dual-task conditions was significantly higher for the visuospatial memory task than for the digit memory task for the longer target duration, t(15) = 2.50, p = .025, but not for the shorter one, t(15) = 1.28, p = .220. For the auditory modality stimulus duration, the ANOVA on the proportion of accurate responses revealed significant main effects of age, F(1, 22) = 229.98, p < .0001, g2p = .91, task condition, F(2, 44) = 19.48, p < .0001, g2p = .47, and memory task, F(1, 22) = 16.44, p = .001, g2p = .43, as well as a significant interaction between task condition and memory task, F(2, 44) = 5.06, p = .01, g2p = .19. No other interaction effect reached statistical significance: Task Condition  Age, F < 1, Memory Task  Age, F(1, 22) = 1.67, p = .210, and Task Condition  Memory Task  Age, F < 1. Post hoc tests indicated that for the visuospatial memory task, the proportion of accurate responses did not change as a function of task condition: single task versus 6-s dual task, t < 1, single task versus 12-s dual task, t(23) = 1.90, p = .07, and 6-s dual task versus 12-s dual task, t < 1. In contrast, for the digit memory task, the proportion of accurate responses was higher in the single-task condition, t(23) = 2.70, p = .013, than in the 6- and 12s dual-task conditions, t(23) = 7.25, p < .0001, and it was also higher in the 6-s dual-task condition than in the 12-s one, t(23) = 3.60, p = .002. Discussion The subjective shortening effect in the dual-task condition observed in Experiment 1 was replicated in the current experiment with two different concurrent nontemporal tasks, namely a digit memory task and a visuospatial memory task. Unexpectedly, this shortening effect was particularly high for the visual 12-s target duration and the visuospatial memory task. This could result from the high processing demands imposed by this specific dual-task condition. First, as explained earlier, tracking the passage of time for a long duration requires more sustained attentional effort. Second, the attention control switch system connecting the pacemaker to the accumulator is maintained with more difficulty in a closed state while perceiving visual stimuli than perceiving auditory stimuli (e.g., Penney, 2003). Put differently, timing visual stimuli is more attentional demanding than timing auditory stimuli. That might also explain why the subjective shortening in the dual-task condition— regardless of the type of memory task used (visuospatial or digit)—was systematically higher for the visual stimuli than for the auditory stimuli. Third, the visuospatial memory task was more difficult than the digit memory task, thereby consuming more attentional resources. Indeed, our results revealed that in both age groups the proportion of accurate responses was lower in the visuospatial memory task than in the digit memory task.

A.-C. Rattat / Journal of Experimental Child Psychology 106 (2010) 145–162

159

In contrast, whereas in Experiment 1 the timing task systematically interfered with the concurrent nontemporal task, the effect of adding a processing of time on nontemporal performance depended on duration modality in Experiment 2. More precisely, regardless of age group, timing an auditory stimulus decreased the proportion of accurate responses in the concurrent digit memory task but not in the concurrent visuospatial memory task. Conversely, timing a visual stimulus decreased the proportion of accurate responses in the concurrent visuospatial memory task in both age groups and also, to a lesser extent, in the concurrent digit memory task solely in the 8-year-olds. Thus, this asymmetric interference pattern suggests that duration reproduction requires not only executive resources but also phonological and/or visuospatial resources as a function of the modality in which the stimulus to be timed was presented.

General discussion The aim of this study was to examine the nature of the resources dedicated to timing in children. The originality of our findings concerns the developmental aspects of bidirectional interference between temporal reproduction and different concurrent nontemporal processing. The main findings from this study can be summarized as follows. First, in both age groups, adding a concurrent memory task had no effect on time reproduction variability, although it did systematically affect time accuracy by producing a shortening effect, especially for the longest target duration (12 s) and especially for the visual modality. Second, and conversely, the timing of visual or auditory stimuli decreased the proportion of accurate responses in the concurrent executive task. The decrease in the proportion of accurate responses under dual-task conditions was nevertheless greater in the 5-year-olds than in the 8-year-olds, especially for the longest target duration. Third, and in contrast, when the concurrent task was the digit or visuospatial memory task, the effect of adding a processing of time on nontemporal performance depended on the modality in which the stimulus was presented. Timing an auditory stimulus decreased the proportion of accurate responses in the digit memory task but not in the visuospatial memory task, whereas timing a visual stimulus decreased the proportion of accurate responses in the visuospatial memory task but not in the digit memory task, at least to a lesser extent in the 8-year-olds. The first finding is in line with the studies described in the Introduction, highlighting the underestimation of durations in dual-task conditions compared with single-task conditions both by children (Gautier & Droit-Volet, 2002) and by adults (e.g., Champagne & Fortin, 2008; Kladopoulos et al., 2004). Within the context of accumulator timing models, this shortening effect observed in a dual-task condition is conventionally attributed to a loss of temporal information (pulses) due to attentional resources being drawn away from timing. More interesting, our data show experimentally for the first time that this shortening effect in a dual-task condition appeared regardless of the sensory modality in which the stimulus is presented (visual or auditory) and regardless of the type of memory task used (executive task or digit or visuospatial memory task). This provides support for claims that temporal processing involves a working memory control system with limited attentional capacity, that is, the central executive (Brown, 1997, 2006; Brown & Frieh, 2000; Fortin et al., 2007; Rammsayer & Ulrich, 2005; Zakay & Block, 2004). Additional evidence for this comes from neuroscience studies showing that working memory and timing rely on the same anatomical structures (Lewis & Miall, 2003, 2006; Lustig, Matell, & Meck, 2005). More specifically, Lewis and Miall (2006) found evidence that the processing of durations of more than 1 s activates the right hemispheric dorsolateral prefrontal cortex brain region strongly associated with central executive functions. The second finding from our study indicates that, regardless of duration modality, concurrent timing interfered with executive performance, thereby providing additional support for the idea that timing involves executive resources. As Brown (2006) put it, ‘‘If two tasks depend on a common set of resources, then one may observe a bidirectional interference. That is, each task interferes with performance on the other” (p. 1464). Our findings contribute to the literature by showing that this bidirectional interference under dual-task conditions was greater in the 5-year-olds than in the 8-year-olds, especially for the longest target duration (12 s). As revealed in Experiment 1, the decrease in the proportion of correct responses in the executive task between the single- and dual-task conditions was greater in the 5-year-olds than in the 8-year-olds for the 12-s target duration. How can we explain

160

A.-C. Rattat / Journal of Experimental Child Psychology 106 (2010) 145–162

these age-related shifts in the effect of timing on simultaneous central executive processing? As explained earlier, there is a general consensus that the capacity of the central executive increases from 5 or 6 years of age to early adolescence (Gathercole, 1998). During temporal processing, therefore, the amount of executive resources left over for the processing of nontemporal information increases with age, especially for longer durations when more resources are required to track the passage of time. Consequently, when the longest duration (12 s) was used, the 5-year-olds had clearly insufficient executive resources available to achieve optimum nontemporal performance in the dual-task condition. The question arises about the extent to which the modality used for the experimental stimuli in the executive task, namely the auditory modality, influenced the pattern of mutual interference observed in our first experiment. As explained in the Introduction, the phonological loop in Baddeley and Hitch’s (1974) working memory model processes auditory verbal material. Thus, our pattern of interference could be ascribed to competition for phonological resources. Insofar as the bidirectional interference effects in our experiment did not change as a function of duration modality, the above assumption implies that the processing of both auditory and visual durations necessarily requires phonological resources. This is not, however, supported by the third finding from our study, which suggests that the involvement of the two slave memory mechanisms depends on the modality in which the stimulus to be timed is presented. Our results in Experiment 2 indeed show that concurrent auditory timing solely disrupts performance in a memory task primarily taxing phonological resources (digit memory task), whereas concurrent visual timing solely disrupts visuospatial memory performance. Note, however, that in the 8-year-olds the concurrent visual timing also interfered with the digit performance, albeit to a lesser extent. Thus, the data from the younger children suggest that the processing of visual stimuli involves the visuospatial sketchpad but not the phonological loop memory system. A further interesting point concerns this age-related change in the involvement of the slave systems in visual timing. Hitch and colleagues (Hitch & Halliday, 1993; Hitch, Halliday, Dodd, & Littler, 1989; Hitch, Halliday, Schaafstal, & Schraagen, 1998) ran experiments on 5- and 7- or 8-year-olds on memory for pictures of objects, that is, visual material that is recordable into a phonological form. Their results suggested that the 5-year-olds, but not the 8-year-olds, are stimulus modality dependent. Thus, although both the phonological loop and visuospatial sketchpad memory systems are present at this age (Gathercole, Pickering, Ambridge, & Wearing, 2004), the younger children process stimulus information with the specific modality-related memory system, that is, the visuospatial sketchpad for visual stimuli (pictures of objects). In contrast, the 8-year-olds adopt a strategy of verbally recoding pictures where possible and so use the phonological loop to mediate memory of the pictures. This verbal recoding would be due to the emergence of subvocal rehearsal strategies at around 7 years of age (Bjorklund & Coyle, 1995; Gathercole & Hitch, 1993; Hulme et al., 1984). Put differently, there would be increasing use by the older children of a nonvisual strategy using the phonological loop (i.e., rehearsal) to supplement the contribution of the visuospatial sketchpad in their memory for visual information (e.g., Miles, Morgan, Milne, & Morris, 1996). That might also explain why in our study the concurrent visual timing interfered with both the visuospatial and digit memory performances in the 8-year-olds, whereas it interfered with only the visuospatial memory performance in the 5-yearolds. One question that arises from this consideration concerns the verbal code used. On the basis of the available data, it is difficult to answer. One possibility is that despite the fact that all children were explicitly told not to count, part of the older age group did nevertheless use a counting strategy. Thus, these 8-year-olds may have used a verbal recoding of visual duration (e.g., a semantic value of the estimated duration) that might partly account for the bidirectional interference between visual timing and concurrent digit memory task observed at this age. Be that as it may, further research is now necessary to investigate this phenomenon of verbal recoding of duration in memory. To conclude, this study is the first to examine the extent to which executive, phonological, and visuospatial resources are involved in duration processing in children. The comparisons between 5- and 8-year-olds provide evidence that temporal processing involves both the central executive and the slave memory system associated with the modality of the temporal stimulus but that the older children are likely to use phonological resources to rehearse visual temporal stimuli.

A.-C. Rattat / Journal of Experimental Child Psychology 106 (2010) 145–162

161

References Arlin, M. (1986a). The effect of quantity, complexity, and attentional demand on children’s time perception. Perception & Psychophysics, 40, 177–182. Arlin, M. (1986b). The effect of quantity and depth of processing on children’s time perception. Journal of Experimental Child Psychology, 42, 84–98. Baddeley, A. D., & Hitch, G. (1974). Working memory. In G. A. Bower (Ed.), Recent advances in learning and motivation (pp. 47–90). New York: Academic Press. Bauer, P. J., & Shore, C. M. (1987). Making a memorable event: Effects of familiarity and organization on young children’s recall of action sequences. Cognitive Development, 2, 327–338. Bjorklund, D. F., & Coyle, T. R. (1995). Utilization deficiencies in the development of memory strategies. In F. E. Weinert & W. Schneider (Eds.), Memory performance and competencies: Issues in growth and development (pp. 161–180). Hillsdale, NJ: Lawrence Erlbaum. Brannon, E. M., Roussel, L. W., Meck, W. H., & Woldorff, M. (2004). Timing in the baby brain. Cognitive Brain Research, 21, 227–233. Brown, S. W. (1997). Attentional resources in timing: Interference effects in concurrent temporal and nontemporal working memory tasks. Perception & Psychophysics, 59, 1118–1140. Brown, S. W. (2006). Timing and executive function: Bidirectional interference between concurrent temporal production and randomization tasks. Memory & Cognition, 34, 1464–1471. Brown, S. W., & Frieh, C. T. (2000). Information processing in the central executive: Effects of concurrent temporal production and memory updating tasks. In P. Desain & L. Windsor (Eds.), Rhythm perception and production (pp. 193–196). Lisse, Netherlands: Swets & Zeitlinger. Brown, S. W., & Merchant, S. M. (2007). Processing resources in timing and sequencing tasks. Perception & Psychophysics, 69, 439–449. Casini, L., & Macar, F. (1997). Effects of attention manipulation on judgments of duration and of intensity in the visual modality. Memory & Cognition, 25, 812–818. Champagne, J., & Fortin, C. (2008). Attention sharing during timing: Modulation by processing demands of an expected stimulus. Perception & Psychophysics, 70, 630–639. Cohen, J., MacWhinney, B., Flatt, M., & Provost, J. (1993). PsyScope: An interactive graphic system for designing and controlling experiments in the psychology laboratory using Macintosh computers. Behavior Research Methods, Instruments, & Computers, 25, 257–271. Cowan, N. (1997). The development of memory in childhood. Hove, UK: Psychology Press. Cowan, N. (2005). Working memory capacity. Hove, UK: Psychology Press. Droit-Volet, S. (2003a). Alerting attention and time perception in children. Journal of Experimental Child Psychology, 85, 372–384. Droit-Volet, S. (2003b). Temporal experience and timing in children. In W. H. Meck (Ed.), Functional and neural mechanisms of interval timing (pp. 183–208). Boca Raton, FL: CRC Press. Droit-Volet, S., Clément, A., & Fayol, M. (2003). Time and number discrimination in a bisection task with a sequence of stimuli: A developmental approach. Journal of Experimental Child Psychology, 84, 63–76. Droit-Volet, S., Delgado, M., & Rattat, A.-C. (2006). The development of the ability to judge time in children. In J. R. Marrow (Ed.), Focus on child psychology research (pp. 81–104). New York: Nova Science. Droit-Volet, S., & Rattat, A.-C. (1999). Are time and action dissociated in young children’s time estimation? Cognitive Development, 14, 573–595. Droit-Volet, S., & Wearden, J. H. (2001). Temporal bisection in children. Journal of Experimental Child Psychology, 55, 142–159. Droit-Volet, S., Wearden, J. H., & Delgado, M. D. (2007). Short-term memory for time in children and adults: A behavioral study and a model. Journal of Experimental Child Psychology, 97, 246–264. Fortin, C., Champagne, J., & Poirier, M. (2007). Temporal order in memory and interval timing: An interference analysis. Acta Psychologica, 126, 18–33. Franssen, V., Vandierendonck, A., & Van Hiel, A. (2006). Duration estimation and the phonological loop: Articulatory suppression and irrelevant sounds. Psychological Research, 70, 304–316. Gathercole, S. E. (1998). The development of memory. Journal of Child Psychology and Psychiatry, 39, 3–27. Gathercole, S. E. (2002). Memory development during the childhood year. In A. D. Baddeley, M. D. Kopelman, & B. A. Wilson (Eds.), Handbook of memory disorders (2nd ed., pp. 475–500). Chichester, UK: John Wiley. Gathercole, S. E., & Hitch, G. J. (1993). Developmental changes in short-term memory: A revised working memory perspective. In A. Collins, S. E. Gathercole, M. A. Conway, & P. E. Morris (Eds.), Theories of memory (pp. 189–210). Hove, UK: Lawrence Erlbaum. Gathercole, S. E., Pickering, S. J., Ambridge, B., & Wearing, H. (2004). The structure of working memory from 4 to 15 years of age. Developmental Psychology, 40, 177–190. Gautier, T., & Droit-Volet, S. (2002). Attention and time estimation in 5- and 8-year-old children: A dual-task procedure. Behavioural Processes, 58, 57–66. Gibbon, J., Church, R. M., & Meck, W. H. (1984). Scalar timing in memory. In J. Gibbon & L. G. Allan (Eds.), Annals of the New York Academy of Sciences: Timing and time perception (pp. 457–461). New York: New York Academy of Sciences. Hitch, G. J., & Halliday, M. S. (1993). Working memory in children. Philosophical Transactions of the Royal Society of London B, 302, 324–340. Hitch, G. J., Halliday, M. S., Dodd, A., & Littler, J. E. (1989). Development of rehearsal in short-term memory: Differences between pictorial and spoken stimuli. British Journal of Developmental Psychology, 7, 347–362. Hitch, G. J., Halliday, M. S., Schaafstal, A. M., & Schraagen, J. M. C. (1998). Visual working memory in young children. Memory & Cognition, 16, 120–132. Hulme, C., Thomson, N., Muir, C., & Lawrence, A. (1984). Speech rate and the development of short-term memory span. Journal of Experimental Child Psychology, 38, 241–253.

162

A.-C. Rattat / Journal of Experimental Child Psychology 106 (2010) 145–162

Imbo, I., Vandierendonck, A., & Vergauwe, E. (2007). The role of working memory in carrying and borrowing. Psychological Research, 71, 467–483. James, W. (1890). The principles of psychology. New York: Henry Holt. Kladopoulos, C. N., Hemmes, N. S., & Brown, B. L. (2004). Prospective timing under dual-task paradigms: Attentional and contextual-change mechanisms. Behavioural Processes, 67, 221–233. Lewis, P. A., & Miall, R. C. (2003). Overview: An image of human neural timing. In W. H. Meck (Ed.), Functional and neural mechanisms of interval timing (pp. 515–532). Boca Raton, FL: CRC Press. Lewis, P. A., & Miall, R. C. (2006). Remembering the time: A continuous clock. Trends in Cognitive Science, 10, 410–416. Lustig, C., Matell, M. S., & Meck, W. H. (2005). Not ‘‘just” a coincidence: Frontal–striatal interactions in working memory and interval timing. Memory, 13, 441–448. Meck, W. H., & Malapani, C. (2004). Neuroimaging of interval timing. Cognitive Brain Research, 21, 133–137. Miles, C., Morgan, M. J., Milne, A. B., & Morris, E. D. M. (1996). Developmental and individual differences in visual memory span. Current Psychology, 15, 53–67. Penney, T. B. (2003). Modality differences in interval timing: Attention, clock speed, and memory. In W. H. Meck (Ed.), Functional and neural mechanisms of interval timing (pp. 209–234). Boca Raton, FL: CRC Press. Predebon, J. (1996). The effects of active and passive processing of interval events on prospective and retrospective time estimates. Acta Psychologica, 94, 41–58. Rammsayer, T., & Ulrich, R. (2005). No evidence for qualitative differences in the processing of short and long temporal intervals. Acta Psychologica, 120, 141–171. Rattat, A.-C., & Droit-Volet, S. (2002). Le transfert d’un apprentissage de durée d’action chez le jeune enfant: L’effet facilitateur de la variété des actions? Enfance, 2, 141–153. Rattat, A.-C., & Droit-Volet, S. (2005). The long-term retention of time: A developmental study. Quarterly Journal of Experimental Psychology B, 58, 163–176. Rattat, A.-C., & Droit-Volet, S. (2007). Implicit long-term memory for duration in young children. European Journal of Cognitive Psychology, 19, 271–285. Riggs, K. J., McTaggart, J., Simpson, A., & Freeman, R. J. (2006). Changes in the capacity of visual working memory in 5- to 10year-olds. Journal of Experimental Child Psychology, 95, 18–26. Sawyer, T. (2003). Allocation of attention and production of time intervals. Perceptual & Motor Skills, 96, 905–908. Szmalec, A., Vandierendonck, A., & Kemps, E. (2005). Response selection involves executive control: Evidence from the selective interference paradigm. Memory & Cognition, 33, 531–541. Tartas, V. (2001). The development of systems of conventional time: A study of the appropriation of temporal locations by fourto ten-year-old children. European Journal of Psychology of Education, 16, 197–208. Towse, J. N., Hitch, G. J., & Horton, N. (2007). Working memory as an interface between processing and retention: A developmental perspective. Advances in Child Development & Behavior, 35, 219–251. Ulbrich, P., Churan, J., Fink, M., & Wittmann, M. (2007). Temporal reproduction: Further evidence for two processes. Acta Psychologica, 125, 51–65. Wearden, J. H. (2005). Origines et développement des théories d’horloge interne du temps psychologique. Psychologie Française, 50, 7–25. Wearden, J. H., & Culpin, V. (1995). Exploring scalar timing theory with human subjects. In M. Richelle, V. De Keyser, G. D’Ydewalle, & A. Vandierendonck (Eds.), Time and the dynamic control of behavior (pp. 21–33). Liège, Belgium: University of Liège Press. Wearden, J. H., & Culpin, V. (1998). Exploring scalar timing theory with human subjects. In V. De Keyser, G. D’Ydewalle, & A. Vandierendonck (Eds.), Time and the dynamic control of behavior (pp. 33–49). Göttingen, Germany: Hogrefe & Huber. Wearden, J. H., Parry, A., & Stamp, L. (2002). Is subjective shortening in human memory unique to time representations? Quarterly Journal of Experimental Psychology B, 55, 1–25. Wilkening, F., Levin, I., & Druyan, S. (1987). Children’s counting strategies for time quantification and integration. Developmental Psychology, 23, 823–831. Zakay, D., & Block, R. A. (2004). Prospective and retrospective duration judgments: An executive-control perspective. Acta Neurobiologiae Experimentalis, 64, 319–328.