Are Time and Action Dissociated in Young Children's Time Estimation?

Cognitive Development, 14, 573–595 (1999) © 2000 Elsevier Science Inc.ISSN 0885-2014 All rights of reproduction reserved.

Manuscript received November 12, 1997; revision accepted August 17, 1999 573

Are Time and Action Dissociated in YoungChildren’s Time Estimation?

Sylvie Droit-VoletAnne-Claire Rattat

Blaise Pascal University, Clermont-Ferrand, France

Experiment 1 examined whether there is a developmental shift in children’sability to differentiate a given amount of time from a particular action. Inthree sessions, 3- and 5 1/2-year-olds were trained to produce an action (i.e.,pressing on a squeezer) for 5 s. Twenty-four hours later, control participantswere required to produce this target duration using the same action, whereasexperimental participants had to do so with a new action (i.e., pressing a but-ton). The results showed that the 5 1/2-year-olds achieved the same temporalperformance in both groups. In contrast, the 3-year-olds’ temporal perfor-mance was significantly better in the control group than in the experimentalgroup. Two additional studies were run with 3-year-olds, the first designedto assess the transfer without a delay, and the second with explicit instruc-tions to transfer duration. In each study, 3-year-olds’ temporal performancewas significantly better in the control group than in the experimental group.These findings as a whole suggest that 3-year-olds fail to understand that oneand the same duration can be shared by several different actions. Early im-plicit knowledge of time was discussed.

Temporal order and duration are two fundamental temporal properties of events.In the field of developmental psychology, only knowledge about temporal orderhas been extensively studied in young children (Bauer & Herstsgaard, 1993;Bauer & Mandler, 1989, 1992; Bauer & Shore, 1987; Bauer & Travis, 1993;O’Connel & Gerard, 1985). The lack of studies on knowledge about duration ismainly due to the strong theoretical position of cognitivists who consider dura-

The authors extend their gratitude to the staff and children of the Ecole Jules Ferry, Clermont-Fer-rand, and to the inspector of the academy without whom this study could not have been carried out.

Direct all correspondence to: Sylvie Droit-Volet, Laboratoire de Psychologie Sociale de laCognition, CNRS (UPRE-SA 6024), Université Blaise Pascal, 34 avenue Carnot, 63007 Cler-

mont-Ferrand Cedex, France; E-mail: [email protected].

574 Droit-Volet and Rattat

tion as a sophisticated product of the mind (Anderson, 1983; Ornstein, 1969). InAnderson’s (1983) temporal string model, duration is inferred from temporal or-der, which is the only temporal information stored in long-term memory.

Most developmentalists have provided data consistent with the cognitivestandpoint (Pouthas, Droit, & Jacquet, 1993). They have shown that the ability tojudge time emerges relatively late in ontogenesis, at about age 5 years. Further-more, they agree that children infer duration from other information. Debate hascentered, however, on the nature of this information. Piaget (1946), who workedon kinetic events, showed that time judgment depends upon the ability to coordi-nate movements with their speed. In a classical Piagetian task, children wereshown two cars which started and stopped running together on parallel tracks, butthe one running faster stopped further ahead. In this condition, up until the acqui-sition of concrete operations, children claim that the car stopping further aheadruns for a longer time. Based on their studies of temporal comparison tasks withstatic events, Levin (1992) and Montangero (1977, 1979, 1993) argued that dura-tion was instead derived from the coordination between beginnings and endingsof events (i.e., temporal order).

However, in these temporal comparison tasks, children’s difficulties in judg-ing time may result from limited general cognitive abilities to coordinate severalcues rather than from a specific problem of time estimation (Levin, 1979; Richie& Bickhard, 1988; Wilkening, Levin, & Druyan, 1981). Thus, in order to exam-ine time estimation in younger children, psychologists have simply asked youngchildren to produce or to reproduce a single duration (Crowder & Hohle, 1970;Droit-Volet, 1999a; Fraisse, 1948; Fraisse & Orsini, 1958; Macar, 1988; Mat-suda & Matsuda, 1983; Montangero, 1977). However, in spite of this task simpli-fication, children still fail to show accurate temporal judgment below the criticalage of 5 years. Therefore, Fraisse (1982) concluded that a prerequisite for accu-rate time judgment is the acquisition, at about 5 years, of an explicit knowledgeabout time, where time is abstracted from events and allows to measure eventswhatever their characteristics.

More recently, psychologists have succeeded in observing temporal judgmentabilities in 3-year-olds, who have not yet mastered the concept of homogeneoustime. Friedman (1990) showed that 3-year-olds were successful in comparing theduration of daily activities by placing a marker on a judgment scale going from“a very short time” to a “very long time,” although their judgment lacked preci-sion. More recently, Droit (1994, 1995) showed that 3-year-olds were able to ac-curately produce the target duration of an action when trained to do so.

Three-year-olds’ temporal ability seems to be specific to the duration ofevents or actions (filled duration). Their performance in estimating temporal in-tervals (empty duration) is far inferior. Indeed, Droit (1994) observed, with aconditioning technique, that 3-year-olds could learn to regulate a button-pressduration, but not a temporal interval between two presses, as shown in previousstudies (Bentall, Lowe, & Beasty, 1985; Droit, Pouthas, & Jacquet, 1990, 1991).

A Dissociation between Time and Action 575

In addition, 3-year-olds required personal experience with the critical durationthrough their action, guided either by a model to imitate (Droit, 1995) or by anexternal clock (Droit, 1994), where they were trained to produce the target dura-tion in synchronization with the experimenters’ action or the lighting of bulbs, re-spectively. Without this kind of previous training, in a single session, their tem-poral performance remained poor, even with filled duration (Droit-Volet, 1999a;Matsuda, Miyazaki, & Matsuda, 1983).

On the basis of these empirical data, Droit (1995, 1998, 1999b) concluded that3-year-olds possess implicit procedural time knowledge that radically differsfrom a mature concept of time. In others words, they know how to time their ac-tions before they know that their actions take time. She proposed that children atthis age possess a kind of action-specific time. More precisely, given that an ac-tion may be automatically encoded with its duration (Michon, 1990), young chil-dren should therefore be able to reactivate duration associated with an initial ac-tion on a retrieval task (e.g., time production on several sessions). This wouldexplain 3-year-olds’ failure at traditional single time production tasks, in whichthey have no opportunity to remember or to experience a target duration throughtheir action. Finally, in these tasks, explicit knowledge about duration would fa-cilitate the processing of time by selectively orienting children’s attention to tem-poral information (Droit, 1997).

Although children’s performance improves in accordance with the temporalfeatures of a situation until age 5, children remain effectively unable to articulatetemporal rules governing their behavior (Bentall, Lowe, & Beasty, 1985; Droit,Pouthas, & Jacquet, 1990; Pouthas, Droit, Jacquet, & Wearden, 1990). In tempo-ral learning, when successful 3-year-olds were asked how they got the positivefeedback (i.e., laughing clown), which was contingent on action duration, theynever gave temporal rules. They only said: “I must have the laughing clown” or“I have to win” (Droit, 1998). Recently, Droit-Volet (1998) showed that 3-year-olds produced more accurate response durations when they were instructed topress “hard enough” (i.e., force rule) rather than to press “long enough” (i.e.,temporal rule). Moreover, when they were instructed to press “harder,” theypressed both harder and longer. In the same situation, older children aged 5 1/2-year-olds, only pressed harder. They never pressed longer. Thus, a force rule, nota time rule, seems to control time productions in young children. This estimationof time via a force rule in 3-year-olds reveals that time is still not dissociatedfrom action (Arlin, 1986, 1989; Levin, 1992; Matsuda, 1991; Montangero,1977). Consequently, a duration initially encoded with an action would not be as-signed to a possible other action. It is only when children begin to differentiatetime from action, at about age 5, that a duration learned with an action should betransferred to another action.

In the present study, the transfer paradigm was used to investigate the natureof 3-year-olds’ filled duration knowledge. The hypothesis was that with implicittime knowledge, 3-year-olds are able to remember a given duration acquired


throughout the practice of an action, but not, unlike 5 1/2-year-olds, to transferthis same duration to a new action. Three- and 5 1/2-year-old participants per-formed a response duration task over four sessions with a target duration of 5 s(

1

/

2

1 s). During the first three sessions, they were trained to produce the targetduration on a responding system (pressing a rubber squeezer) by imitating an ex-perimenter (i.e., simultaneous imitation). In the last session, they were randomlyassigned either to an experimental group, involving a different response system(pressing a button), or to a control group involving the same response system asthat used during the training sessions. We expected that the 3-year-olds wouldproduce more accurate response durations in the control condition than in the ex-perimental condition, whereas the 5 1/2-year-olds would produce the same levelof accurate response durations in the control condition as in the experimentalcondition. Accordingly, the 3-year-olds should produce shorter and more vari-able durations in the experimental group than in the control group, whereas nodifference should emerge between the two groups of 5 1/2-year-olds.

E

XPERIMENT

1

Method

Participants.

The final sample consisted of 40 children

1

: twenty 3-year-olds(10 girls and 10 boys, mean age

5

3.3 years;

SD

5

0.2) and twenty 5 1/2-year-olds (10 girls and 10 boys, mean age

5

5.6 years;

SD

5

0.3). They were re-cruited from the same nursery school in Clermont-Ferrand, France.

Materials.

The child was seated at a table opposite a toy theater and next tothe experimenter. Two response systems were used. The first response systemwas a red rubber squeezer, such as the one used in Macar’s (1988) study (diame-ter 5 cm). The child had to press the ball in his or her hand. The second responsesystem was a red button mounted on a wooden box. The two response systemswere connected to a microcomputer which stored each press duration and gener-ated visual feedback. The visual feedback was the projection of a slide for 3 s onthe toy theater’s screen. When the child produced an accurate response, the slidewas a colored picture taken from a popular cartoon. When the child produced aninaccurate response, the slide was black. Slides were grouped in series defined bythe type of cartoon: Snow White, The Lion King, Merlin, etc. The cartoon wasdifferent for each experimental session.

1

Six additional children (four 3-year-olds and two 5 1/2-year-olds) participated in the experiment,but they were not given the last transfer test because they did not produce a sufficient number ofaccurate responses after imitation training.


Procedure.

The child’s task was to produce a target response duration of 5 son six phases of 20 trials each distributed on four sessions, one per day as de-scribed in Table 1. When the child produced a response duration between 4 and 6s, the feedback was a colored slide. When the response duration was less than 4 sor more than or equal to 6 s, the given feedback was the black slide. If the dura-tion reached 10 s, the black slide was presented directly.

In the first three sessions, each participant received an imitation training phasewith the same response system (i.e., a rubber squeezer; Table 1). In the imitationphase, the experimenter had her own rubber squeezer, similar to the child’s. Atthe beginning of the first session, she told the child to imitate her while saying,“Look, I’m going to teach you how to make cartoon pictures appear. Do exactlyas I do.” For the first trial only, the experimenter pressed with her right hand onher own response squeezer during 5 s while pressing the child’s hand on his orher squeezer with her left hand. Then, she released simultaneously her ownsqueezer and her left hand on the child’s. The color slide appeared and the exper-imenter added, “We’ve made it, we’ve got a cartoon picture! If we hadn’t done itright, we would have got a black picture.” Finally she said, “Now, let’s try to-gether to get a cartoon picture. Ready, let’s do exactly as I do.” Before each trial,the experimenter repeated, “Ready, let’s do exactly as I do” and made sure thatthe child synchronized his or her press with hers.

On session 1, immediately before the first imitation phase, the spontaneous re-sponse durations produced with the rubber squeezer were collected in a baselinephase. The experimenter introduced this pretraining phase by saying, “Take thesqueezer in your hand, and play at making cartoon pictures appear.”

To evaluate the effect of imitation training on participants’ response dura-tions, children were submitted to a retention test given immediately after the lastimitation phase in session 3. The experimenter introduced this phase by saying,“Now, you try to make cartoon pictures appear all by yourself!” In this phase, theresponse system was still the rubber squeezer and the child was not given simul-taneous imitation instructions. Four 3-year-olds and two 5 1/2-year-olds did not

Table 1. Experiment 1: The Response System (Squeezer or Button) Used in the Different Phases of 20 Trials of Each Experimental Session

Control Group Experimental Group

Session 1: day 1Baseline Phase Squeezer SqueezerImitation Phase Squeezer Squeezer

Session 2: day 2Imitation Phase Squeezer Squeezer

Session 3: day 3Imitation Phase Squeezer SqueezerRetention Phase Squeezer Squeezer

Session 4: day 4Transfer Test Squeezer Button


participate in the transfer test, because they did not learn the target response dura-tion. Two criteria of learning were chosen on the basis of the oldest children’sperformance on a similar temporal production task with imitation instructions(Droit, 1995): (a) 50% of response durations between 3 and 7 s of which at least35% were between 4 and 6 s, and (b) a modal response distribution focused onthe target duration (5 s). The six participants who did not meet these criteria werenot included in the final sample.

The transfer test was administered in a fourth session. For this test, the 3- andthe 5 1/2-year-olds were randomly assigned either to a control group or to an ex-perimental group. In the control group, the child had the same squeezer as in theprevious imitation sessions. In the experimental group, the child used the button.Nothing else was changed in the experimental situation, except that the experi-menter did not have her own response system. In each group, the experimentergave the same instructions at the beginning of the test, “Now, you try makingcartoon pictures appear.” Like the imitation sessions, the test was composed of20 trials.

Results

Training Analysis.

For each experimental phase, the percentage of accurateresponses (i.e., greater than or equal to 4 s and less than 6 s) out of the total num-ber of responses (i.e., 20) was used as an index of temporal differentiation. Fig-ure 1 presents the percentage of accurate responses obtained by the 3- and the5 1/2-year-olds in the control and the experimental groups for the various phasesof the experiment. To ensure that children had achieved an equal level of perfor-mance in the different groups after the training sessions with simultaneous imita-tion, an Age

3

Group

3

Sex analysis of variance (ANOVA) was run on the per-centage of accurate responses (using arcsin transformed data)

2

in the immediateretention phase. There were no main effects of sex,

F

(1, 32)

5

0.12,

p

.

.05, age,

F

(1, 32)

5

2.14,

p

.

.05, or group,

F

(1, 32)

5

1.12,

p

.

.05. The Age

3

Groupinteraction was not significant,

F

(1, 32)

5

0.07,

p

.

.05.Similar 3-way ANOVAs were computed on (a) the mean response duration

and (b) the coefficient of variation,

3

in the immediate retention phase. Theseanalyses revealed neither significant main effects nor significant interactions.

The global pattern of results confirms that the 3- and the 5 1/2-year-olds didnot reach different levels of performance in the control and the experimentalgroups through the training with simultaneous imitation.

Transfer Analysis.

Accurate responses.

A within-subject factor was intro-duced to assess learning effects across the transfer test (Figure 2). The 20 trials

2

In Experiments 1, 2, and 3, for each statistical analysis, the percentages of accurate responseswere subjected to an arcsin transformation. However, Figures report the raw percentages.

3

To assess variability in time production, a coefficient of variation was calculated by dividing thestandard deviation by the mean duration.


Fig

ure

1.E

xper

imen

t 1:

Per

cent

age

of a

ccur

ate

resp

onse

s fo

r 3-

and

5 1

/2-y

ear-

olds

in t

he c

ontr

ol a

nd t

he e

xper

imen

tal g

roup

s as

afu

ncti

on o

f th

e di

ffer

ent

expe

rim

enta

l pha

ses:

(a)

bas

elin

e, (

b) im

med

iate

ret

enti

on, (

c) t

rans

fer

test

.


were grouped into four blocks of five trials. An ANOVA was performed on thepercentage of accurate responses with 3 between-subject factors (Sex, Age,Group) and 1 within-subject factor (Block).

No main effect of sex was observed,

F

(1, 32)

5

0.76,

p

.

.05. The main ef-fects of age,

F

(1, 32)

5

43.36,

p

,

.05, group,

F

(1, 32)

5

4.83,

p

,

.05, andblock,

F

(3, 96)

5

5.67,

p

,

.05, reached significance. The ANOVA also re-vealed significant interactions between age and group,

F

(1, 32)

5

9.92,

p

,

.05,and group and block,

F

(3, 32)

5

4.01,

p

,

.05. The interaction between age andblock was not significant. However, an Age

3

Group

3

Block interactionemerged,

F

(3, 32)

5

4.70,

p

,

.05.A Tukey post-hoc test showed that in each block of trials the 3-year-olds pro-

duced fewer correct responses in the experimental group than in the controlgroup (

p

,

.05). However the experimental participants’ accurate responses werehigher in the transfer test than in the baseline, as indicated by a Wilcoxonmatched-pairs sign ranks test (8.5% vs. 1.5%;

Z

5

2

2,03,

p , .05).4 In contrast,

4 Because the assumption of homogeneity of variance was not met for the baseline phase due tothe high number of participants who did not obtain accurate responses in the baseline phase, anonparametric statistic analysis was run (i.e., the Wilcoxon matched-pairs signed-ranks test).

Figure 2. Experiment 1: Percentage of accurate responses for 3- and 5 1/2-year-olds in the control and the experimental groups as a function of trial blocks in thetransfer test.


for 5 1/2-year-olds, no difference emerged between the experimental group andthe control group, except for trial block 4 in which the experimental participantsproduced more correct responses than the control participants (82% vs. 62%;p , .05).

In addition, a Sheffé post-hoc test revealed that, in the control condition, thepercentage of accurate responses was higher for 5 1/2-year-olds than for 3-year-olds (43% vs. 63%; p , .05). In fact, for 5-year-olds, the percentage of accurateresponses significantly increased between the immediate retention phase and thetransfer test (t-test for paired samples, t(9) 5 22.94, p , .05). For 3-year-olds,the decrease of accurate responses between the immediate retention phase and thetransfer test was not significant (t-test for paired samples, t(9) 5 1.90, p . .05).

Mean response duration and its variability. Age 3 Group 3 Sex 3 BlockANOVAs were computed on the mean response duration and its variability(Table 2).

The ANOVA performed on mean duration showed that the main effect of sexwas not significant, F(1, 32) 5 0.44, p . .05. The main effects of age, F(1, 32) 51.84, p . .05, and block, F(3, 96) 5 0.33, p . .05, were also not significant.Only a main effect of group appeared, F(1, 32) 5 6.13, p , .05. However, agesignificantly interacted with group, F(1, 32) 5 8.04, p , .05. There were no ad-ditional significant interactions.

Planned comparisons showed that the 5 1/2-year-olds produced no differentmean durations in the experimental group than in the control group (5.12 s vs.4.93 s), F(1, 16) 5 0.41, p . .05. In contrast, the 3-year-olds produced shortermean durations in the experimental group than in the control group (3.41 s vs.6.20 s), F(1, 36) 5 7.65, p , .05.

The statistical analysis of the coefficient of variation revealed a significantmain effect of age, F(1, 32) 5 16.78, p , .05. There were no main effects of sex,F(1, 32) 5 0.22, p . .05, group, F(1, 32) 5 0.49, p . .05, or block, F(1, 32) 5

Table 2. Experiment 1: Mean Durations (s) and Coefficients of Variation for 3- and 5 1/2-year-olds Across Trial Blocks in the Control Group and the Experimental Group


M. Coef. M. Coef.

3-year-oldsBlock 1 6.00 .26 3.03 .43Block 2 6.31 .32 3.40 .35Block 3 6.18 .26 3.47 .42Block 4 6.30 .28 3.74 .34

5 1/2-year-oldsBlock 1 5.02 .20 5.26 .19Block 2 4.75 .22 5.36 .00Block 3 4.95 .27 4.88 .15Block 4 5.02 .21 4.98 .11


1.15, p . .05. Age and Group interacted significantly, F(1, 32) 5 5.55, p , .05,but no other significant interactions were found. Finally, planned comparisonsindicated that the 3-year-olds were more temporally variable than the 5 1/2-year-olds in the experimental condition, F(1, 16) 5 14.66, p , .05, but not in the con-trol condition, F(1, 16) 5 2.60, p . .05.

Discussion

As predicted, the 5 1/2-year-olds were able to transfer a target duration fromone action to another. Indeed, their temporal performance (i.e., accurate re-sponses, mean duration, coefficient of variation) was similar in the experimentalcondition and in the control condition. In contrast, the 3-year-olds failed to repro-duce the target duration with a new action. They produced less accurate re-sponses in the experimental group than in the control group. Furthermore, dura-tions were shorter and more variable in the experimental group.

From the accurate transfer in the 5 1/2-year-old children, it can be inferredthat 5 1/2-year-olds have abstracted the duration from the action learned by imi-tation and then assigned this duration to another action. These findings suggestthat, at this age, children understand that the same duration can be shared by twodifferent actions.

Unlike 5 1/2-year-olds, the 3-year-olds’ failure to transfer suggests that dura-tion is still not dissociated from its original action. However, in the present study,definitive conclusions cannot be drawn. Indeed, variables other than children’sunderstanding of time may have played a role preventing them from transferringduration. In the control condition, 3-year-olds were able to remember an action-specific duration after a delay of 24 hours between the retention phase and thetransfer test. But in the experimental condition, the change of surface features(i.e., the use of a response button) may have interfered with the accessibility ofthe critical duration in memory. Thus, a second experiment was conducted to seewhether the 3-year-olds’ failure to transfer persists without a delay of 24 hoursbetween the immediate retention phase and the transfer test.

EXPERIMENT 2

Experiment 2 replicated Experiment 1, except that the transfer test occurred justfollowing the immediate retention phase.

Method

Participants. A total of twenty 3-year-olds (7 girls and 13 boys, mean age 53.02; SD 5 0.30), from the same nursery school in Clermont-Ferrand, partici-pated in this experiment. No children who participated in Experiment 1 partici-pated in this experiment. Four additional children were not retained in the finalsample because they did not reach the training criterion defined above.


Material and Procedure. The apparatus was the same as that used in Exper-iment 1. The procedure was also identical, except for the temporal interval be-tween the immediate retention phase and the transfer test. Thus, in this study,participants were given three successive experimental phases in session 3: (a) si-multaneous imitation, (b) immediate retention, and (c) transfer test.

Results

Training Analysis. Figure 3 shows the percentage of accurate responses(i.e., responses between 4 and 6 s) in the control and the experimental groups as afunction of the three experimental phases: (a) baseline, (b) immediate retention,and (c) transfer test. Group 3 Sex ANOVAs on the three dependent measures(percentage of accurate responses, mean duration and coefficient of variation) inthe immediate retention phase revealed no significant effects (Accurate re-sponses: Sex, F (1, 16) 5 0.55, p . .05, Group, F (1, 16) 5 1.93, p . .05; Mean du-ration: Sex, F (1, 16) 5 0.08, p . .05, Group, F (1, 16) 5 2.56, p . .05; Coefficientof variation: Sex, F (1, 16) 5 0.08, p . .05, Group, F (1, 16) 5 0.28, p . .05).

Transfer Analysis. Accurate responses. For the transfer test, an ANOVAwas carried out on the percentage of accurate responses with 2 between-subjectfactors (Sex, Group) and 1 within-subject factor (Block; Figure 4). The analysisrevealed main effects of group, F (1, 16) 5 31.41, p , .05, block, F(3, 48) 57.13, p , .05, and an Group 3 Block interaction, F (3, 48) 5 6.34, p , .05. Asindicated by a paired-samples t-test, the percentage of accurate responses in the

Figure 3. Experiment 2: Percentage of accurate responses for 3- and 5 1/2-year-olds in the control and the experimental groups as a function of the different experi-mental phases: (a) baseline, (b) immediate retention, (c) transfer test.


control group significantly decreased between block 1 and block 2 (t(9) 53.35, p , .05). There was no main effect of sex.

Mean response duration and its variability. Table 3 shows the mean responseduration and the coefficient of variation in the transfer test for the control and theexperimental 3-year-olds as a function of trial blocks. An ANOVA was com-puted for each measure with sex and group as between-subject factor and blockas within-subject factor. For the mean response duration, a main effect of group,F (1, 16) 5 114.32, p , .05, and an interaction Group 3 Block, F (1, 16) 5 4.09,p , .05, were found. No additional significant effects were observed. For the co-efficient of variation, only a significant Group 3 Block interaction appeared,F (3, 48) 5 2.84, p , .05.

Discussion

The results of Experiment 2 are consistent with those of Experiment 1. Indeed,3-year-olds produced more accurate responses in the control group than in theexperimental group. The response duration was also longer and less variable.However, in the present experiment, an interaction with the trial blocks emerged,resulting from a decrease in control participants’ performance in the second

Figure 4. Experiment 2: Percentage of accurate responses for 3- and 5 1/2-year-olds in the control and the experimental groups as a function of trial blocks in thetransfer test.


block of trials in the transfer test. Session 3, composed of three successive phasesof 20 trials, appeared to be too long for young children. However, this result doesnot throw into question the primary findings.

Thus, consistent with our predictions in Experiment 2 as in Experiment 1, the3-year-olds did not succeed in transferring the critical duration acquired with anaction to a new action. However, in both experiments, children were not explic-itly instructed to perform this duration transfer. Thus, their failure may be due notto the implicit nature of their knowledge, but to a misunderstanding of the taskdemands. Until age 5, children are particularly dependent upon the support pro-vided by the discourse of adults both in the understanding of a specific task andin the direction of their memory search (Bauer, 1997). A third experiment wastherefore conducted to investigate whether the lack of duration transfer in 3-year-olds persists even with explicit instructions pointed out that the response durationrequired for the new response system (i.e., button) is the same as for the previousresponse system (i.e., squeezer).

EXPERIMENT 3

Experiment 2 was replicated in Experiment 3, except that 3-year-olds were nowgiven temporal instructions, as described below.

Method

Participants. Twenty 3-year-olds (11 girls and 9 boys, mean age 5 3.16;SD 5 0.32) participated in the study. One child was not retained in the final sam-ple because he did not reach the training criterion. All children were recruitedfrom the same nursery school in Clermont-Ferrand. No children who participatedin the previous experiments participated in this experiment.

Apparatus and Procedure. The apparatus was the same as that used in Ex-periments 1 and 2. The procedure used was also identical to that used in Experi-ment 2, except that participants received temporal instructions. Indeed, childrenwere given explicit temporal instructions at the beginning of each experimental

Table 3. Experiment 2: Mean Durations (s) and Coefficients of Variation for 3-year-olds Across Trial Blocks in the Control Group and the Experimental Group


M. Coef. M. Coef.

3-year-oldsBlock 1 4.42 0.36 1.04 0.57Block 2 5.48 0.57 0.40 0.33Block 3 5.72 0.75 0.43 0.24Block 4 5.32 0.72 0.55 0.27


phase. The additional temporal instructions were for: (a) the baseline phase: “ . . .to make picture appear, you must press the button/the squeezer long enough;” (b)the copy-imitation phase: . . . “do exactly as I do. We must press long enough onthe button/squeezer;” (c) the immediate retention phase: “now try making thepictures appear all by yourself. Press long enough.” In addition, before the trans-fer test, the experimenter pointed out that the temporal task was the same than inthe immediate retention phase. In the control group, she said, “Now, you try al-ways making cartoon pictures appear, and to do so you must press always thesame time,” and in the experimental group: “Now, we play with a button and notany more with a squeezer. But, the game is the same. You try always making car-toon pictures appear, and to do so you must press always the same time, the sametime with the button than with the sqeezer.”

Results

Training Analysis. As in the previous experiments, ANOVAs computed ondata from the immediate retention phase revealed neither main effects of the ma-nipulated variables nor any interactions (Figure 5) (Percentage of accurate re-sponses (arcsin transformed): Group, F (1, 16) 5 0.09, p . .05, Sex, F (1, 16) 50.001, p . .05, Group 3 Sex, F (1, 16) 5 0.22, p . .05; Mean duration: Group,F (1, 16) 5 1.72, p . .05, Sex, F (1, 16) 5 2.41, p . .05, Group 3 Sex, F (1, 16) 50.64, p . .05; Coefficient of variation: Group, F (1, 16) 5 0.001, p . .05, Sex, F(1, 16) 5 2.67, p . .05, Group 3 Sex, F (1, 16) 5 1.37, p . .05).

Transfer Analysis. Accurate responses. An ANOVA was run on the per-centage of accurate responses (using arcsin transformed data) in the transfer testwith Sex and Group as between-subject factors and Block as within-subject fac-tor (Figure 6). Only a main effect of group, F (1, 16) 5 10.59, p , .05, was ob-served. However, the Wilcoxon matched-pairs signed-ranks test indicated thatthe experimental participants produced more correct responses in the transfer testthan in the baseline (Z 5 2,67, p , .05).

To test the effect of temporal instructions, a combined ANOVA for Experi-ments 2 and 3 was conducted on the percentage of accurate responses with 3 be-tween-subject factors (Experiment, Group, Sex) and 1 within-subject factor(Block). There was neither a main effect of experiment nor an Experiment 3Group interaction, nor an Experiment 3 Group 3 Block interaction. However,experiment interacted significantly with block, F (3, 96) 5 3.37, p , .05. Theanalysis showed main effects of group, F (1, 32) 5 44.62, p , .05, and block, F(3, 96) 5 6.63, p , .05, and Group 3 Block interaction, F (3, 96) 5 7.10, p , .05.

Mean response duration and its variability. ANOVAs were conducted onmean response duration and the coefficient of variation with 2 between-subjectfactors (Group, Sex) and 1 within-subject factor (Block) design (Table 4). TheANOVA on mean response duration revealed neither main effects nor interactions.


Figure 5. Experiment 3: Percentage of accurate responses for 3- and 5 1/2-year-olds in the control and the experimental groups as a function of the different experi-mental phases: (a) baseline, (b) immediate retention, (c) transfer test.

Figure 6. Experiment 3: Percentage of accurate responses for 3- and 5 1/2-year-olds inthe control and the experimental groups as a function of trial blocks in the transfer test.


For the coefficient of variation, only a main effect of group appeared, F (1, 16) 5 4.49,p , .05, indicating that response durations were more variable in the experimentalgroup than in the control group (.53 vs. 46).

For the transfer test, combined ANOVAs were run to compare Experiments 2and 3 (i.e., with and without explicit temporal instructions), with experiment,group, and sex as between-subject factors, and block as within-subject factor. Formean response duration, the ANOVA showed no main effect of experiment.However, experiment interacted significantly with group, F(1, 32) 5 14.33, p ,.05, and subsumed a main effect of group, F (1, 32) 5 25.74, p , .05. No addi-tional significant effects were observed. A Sheffé post-hoc test indicated that ex-perimental participants produced longer durations in Experiment 3 than in Exper-iment 2 (3.06 s vs. 0.60 s, p , .05). For the coefficient of variation, the statisticalanalysis revealed a significant Experiment 3 Group 3 Block interaction, F (3,96) 5 3.14, p , .05. In block 3, experimental participants’ temporal productionwas more variable with than without instructions (Tukey post-hoc test, p , .05).There were no other significant effects.

Discussion

In the present study, as in Experiments 1 and 2, the results showed that 3-year-olds obtained significantly more accurate responses in the control group than inthe experimental group and their temporal productions were less variable. How-ever, contrary to Experiments 1 and 2, 3-year-olds did not produce longer dura-tions in the control group than in the experimental group. These findings suggestthat 3-year-olds pressed longer with temporal instructions but they did not suc-ceed in transferring a given duration from one action to another.

GENERAL DISCUSSION

The findings from these three experiments were highly consistent. They showedthat, without delay between the retention test and the transfer test (Experiment 2),

Table 4. Experiment 3: Mean Durations (s) and Coefficients of Variation for 3-year-olds Across Trial Blocks in the Control Group and the Experimental Group


M. Coef. M. Coef.

3-year-oldsBlock 1 3.75 0.43 3.08 0.50Block 2 3.71 0.46 3.62 0.53Block 3 4.20 0.45 2.92 0.63Block 4 4.28 0.49 2.61 0.43


and with explicit instructions to transfer duration (Experiment 3), 3-year-olds’temporal performance was always better in the control group than in the experi-mental group: these children produced more accurate responses, and their tempo-ral productions were less variable. In contrast, in Experiment 1, 5 1/2-year-olds’temporal performance was similar in the control group and in the experimentalgroup.

In the transfer task, the objection could be made that 3-year-olds failed to ac-cess the temporal rule set up during training. As Zelazo, Frye, and Rapus (1996)have claimed, knowing the rules is not always enough to ensure their application.In Zelazo et al.’s (1996) Cognitive Complexity and Control theory (CCC theory),behavioral changes between 3 and 5 years of age are explained by the acquisitionof increasingly complex rule systems, where complexity corresponds to the num-ber of embedded levels inherent in these systems. Three-year-olds still lackhigher-order rules allowing them to determine which dimension is relevant andthen to select which specific embedded rule to apply. However, as suggested byZelazo et al. (1996), 3-year-olds will consequently persist in applying the rulesthat are strongly associated with the task. In fact, the salient common surface fea-tures (i.e., experimenter, feedback, etc.) in the current experiments, and the struc-tural similarity between the immediate retention phase and the transfer test,should have enhanced children’s tendency to use the established temporal rule. Inaddition, explicit instructions to transfer duration in Experiment 3 might be ex-pected to help children access this rule. In spite of these different supports, how-ever, the temporal transfer never occurred in 3-year-olds, suggesting that 3-year-olds’ difficulty cannot be attributed to difficulty selecting an appropriate explicit rule.

It is nevertheless possible that a single change in surface features, (i.e., rubbersqueezer vs. button) totally disrupts the temporal transfer in young children. Chi,Feltovich, and Glaser (1981) have shown that novice problem solvers in the do-main of physics are often unaware of the structural features of problems andhence treat superficially similar problems as if they were structurally similar.Young children are also known to be particularly sensitive to contextual changes(Nelson, 1986). A new response system perhaps changed the representation ofthe task demands among the present participants. But, here again, the temporalinstructions should have prevented their failure to understand the task demand. Inaddition, the relearning of the target duration in the transfer test should have beenfacilitated. But, in Experiment 1, the results indicated no effect of trial block. Aneffect of block appeared in Experiments 2 and 3, but it resulted from a decreasein performance after several trial blocks. Most likely this indicates that the3-year-olds were tired. Thus, in the end, the results as a whole suggest that the3-year-olds failed to transfer the duration from action to another, because theydid not understand that one and the same duration can be shared by several actions.

Consistent with the present approach, these studies document differences inknowledge about duration for 3- and 5 1/2-year-olds. At age 5 years, the conceptof time has arguably been raised to the status of a homogeneous time abstracted


from events, that can serve to measure two different actions, whatever their char-acteristics. A single duration can therefore be transferred from an action to an-other. At this age, time knowledge has reached a sufficient level of abstraction tobe applied in a wide variety of actions or situations, even in those where theyhave not experienced the duration, as shown by their success in most temporalestimation tasks (Droit-Volet, 1999b).

Although 3-year-olds apparently did not possess explicit knowledge of dura-tion, they were nevertheless capable of remembering the duration of a specificaction after the synchronization training, that is, in a retention test occurring im-mediately (Experiments 2, 3) or 24 hours (Experiment 1) later. These findingssuggest that they constructed implicit procedural knowledge about the durationof an action experienced during training. In contrast with explicit knowledgeabout homogeneous time developed somewhat later, Droit (1995, 1998) andDroit-Volet (1998, 1999b) proposed that 3-year-olds represent a kind of broken-time specific to each action or event, that requires no conscious awareness oftemporal properties of action. Therefore, through practice, the duration would beautomatically encoded with its original action and remain associated with it. Thatwould explain why 3-year-olds were unable to attribute a given duration acquiredwith an action to another action. As suggested by Logan (1988), one of the mainproperties of automaticity is that it is specific to the stimulus and situation experi-enced during training, and typically transfer to new stimuli is very poor.

This idea of an early implicit knowledge of duration encourages a reconsider-ation of the attention models of adult time estimation (Thomas & Weaver, 1975;Zakay, 1989). According to these models, duration is encoded only if attentionresources are allocated to temporal information. In retrospective time estimation,participants are informed that they have to judge the duration of a task when it isfinished. In this condition, with lack of attention, the duration is supposedly notencoded. Temporal judgment is therefore reconstructed from information that re-mains in memory, that is, non-temporal information (Zakay & Block, 1996,1998). For Poynter or Zakay, subjective duration is biased by the amount of seg-ments within an event (Poynter, 1983; Zakay, Tsal, Moses, & Shahar, 1994). Re-cently, Boltz, Kupperman, and Dunne (1998) showed that, even in retrospectiveduration, time judgments become more accurate with experience: the more amelody is familiar, the more accurate is its retrospective estimate. This providesevidence that people can estimate duration without intending to. Thus, there is anincidental learning of duration experienced throughout an action during training.In this theoretical framework, without intending to do so, young children wouldbe able to accurately estimate duration of actions that they have experienced.And their failure in traditional time estimation tasks would result from the factthat the estimates concerned an event that they have not personally experiencedand encoded (implicitly) in memory.

In the developmental psychology of time, memory for duration per se has


been as a whole neglected. However, the present study indicates that after oneday of retention, the 5 1/2-year-old control participants produced more accurateresponses than the 3-year-old controls in the transfer test. Some researchers havestudied how response delay between perceived duration and its judgment affectsthe memory of time (Hawkes, Warshan, & Ray, 1973; Zakay & Fallach, 1984).They showed that the longer the response delay, the greater the difference be-tween objective and subjective duration. Finally, the memory traces about dura-tion could be stronger in 5 1/2-year-olds than in 3-year-olds. Thus, it is possiblethat explicit knowledge of time influences memory of time. Consistent with Sey-mour’s (1980) theory, an additional explicit semantic code, such as “long time”or “short time,” could be associated to the temporal label of actions or events.

One question that should be raised here is whether explicit time knowledgederives from earlier implicit time knowledge. In a Piagetian approach, Karmiloff-Smith (1992, 1994) outlined a process of re-description of implicit knowledgeinto an explicit format. -Volet (1998) showed that 3-year-olds produced more ac-curate response duration with force instructions than with temporal instructions,whereas the opposite was observed in 5-year-olds. According to Droit-Volet(1998), these findings reflect the first stage of the long process of time conceptu-alization. Indeed, the idea of time would derive from an early procedural knowl-edge by way of a feeling of effort. As she wrote, “children’s feeling that some-thing is resisting them through their action may be the first step toward theunderstanding of duration” (p. 247).

However, Karmiloff-Smith (1992, 1994) asserted that this re-description ofknowledge occurs at different ages depending on children’s experience in thedomain of knowledge considered. But, in the time knowledge domain, this re-description occurs in an age range where cognitive shifts have often been putforward in other knowledge domains (Mounoud, 1994; Siegler, 1978; Piaget,1974a, 1974b; Zelazo et al. 1996). Between 3 and 5 years, Piaget (1974a, 1974b)described the passage from a sensori-motor success stage to an understandingstage associated to the awareness of properties of actions. Siegler (1978) re-ported that 3-year-olds seem to proceed in an intuitive manner whereas 5-year-olds adopt analytic systematic strategies. From another perspective, Zelazo et al.(1996) evoked at these ages the growth of a higher-order rule system. Taken to-gether, these findings seem to indicate a general cognitive development of repre-sentational flexibility between 3- and 5-year-olds, that allows children to resist inter-ference from prepotent responses tendencies (Dempster & Brainerd, 1995).

Although it is difficult here to dissociate what was specific to time knowledgeand general to cognitive development, the original procedure of temporal trans-fer, used in the present study, provided results that support our assumption aboutan implicit filled-duration knowledge in 3-year-olds, that differs from the laterexplicit knowledge of time. The problem is now to clarify the relation betweenthese two kinds of knowledge.


REFERENCES

Anderson, J. R. (1983). The architecture of cognition. Cambridge, MA: Harvard Univer-sity Press.

Arlin, M. (1986). The effects of quantity, complexity and attentional demand on chil-dren’s time perception. Perception and Psychophysics, 40, 177–182.

Arlin, M. (1989). The effects of physical work, mental work, and quantity on children’stime perception. Perception and Psychophysics, 45, 209–214.

Bauer, P. (1997). Development of memory in early childhood. In N. Cowan (Ed.), Devel-opment of memory in childhood (pp.83–110). United Kingdom: Psychology Press.

Bauer, P., & Herstsgaard, L. (1993). Increasing steps in recall of events: Factors facilitat-ing immediate and long term memory. Child Development, 64, 1204–1223.

Bauer, P., & Mandler, J. M. (1989). One thing follows another: Effects of temporalstructure on 1- to 2-year-olds’ recall of events. Developmental Psychology, 25,197–206.

Bauer, P., & Mandler, J. (1992). Putting the horse before the cart: The use of temporal or-der in recall of events by one-year-old children. Developmental Psychology, 24, 3,441–452.

Bauer, P., & Shore, C. (1987). Making a memorable event: Effects of familiarity and orga-nization in young children’s recall of actions sequences. Cognitive Development, 2,327–338.

Bauer, P., & Travis, L. (1993). The fabric of an event: Different sources of temporal invariancedifferentially affect 24-month-olds’ recall. Cognitive Development, 8, 319–341.

Bentall, R. P., Lowe, C. F., & Beasty, A. (1985). The role of verbal behavior in humanlearning: II Developmental differences. Journal of the Experimental Analysis of Be-havior, 43, 165–181.

Boltz, M., Kupperman, C., & Dunne, J. (1998). The role of learning in remembered dura-tion. Memory and Cognition, 26, 5, 903–921.

Chi, M., Feltovich, P., & Glaser, R. (1981). Categorization and representation of physicsproblems by experts and novices. Cognitive Science, 5, 121–152.

Crowder, A., & Hohle, R. (1970). Time estimation by young children with and without in-formational feedback. Journal of Experimental Child Psychology, 10, 295–307.

Dempster, F. N., & Brainerd, C. J. (1995). Interference and inhibition in cognition. NewYork: Academic Press.

Droit, S. (1994). Temporal regulation of behavior with an external clock in 3-year-oldchildren: Differences between waiting and response duration tasks. Journal of Ex-perimental Child Psychology, 58, 332–345.

Droit, S. (1995). Learning by doing in 3- and 4-year-old children: Adapting to time. Euro-pean Bulletin of Cognitive Psychology, 14, 283–299.

Droit, S. (1997). Focaliser l’attention de l’enfant sur l’information temporelle: un préreq-uis pour l’estimation du temps. In D. Mellier & A. Von Hofe (Eds.), Attention etcontrôle cognitif : mécanisme, développement des habiletés, pathologies (pp. 185–228). Rouen: Presses Universitaires de Rouen.

Droit, S. (1998). Adaptation to time in preschool children. In V. De Keyser, G. D’Yde-walle, & A. Vandierendonck (Eds.), Time and the Dynamic Control of Behavior(pp. 157–177). Seattle: Hogrefe & Huber Publishers.


Droit-Volet, S. (1998). Adaptation to time in young children: An initial force rule govern-ing temporal behavior. Journal of Experimental Child Psychology, 68, 236–249.

Droit-Volet, S. (1999a). Time estimation in young children: Effects of response type andfamiliarity. Current Psychology of Cognition, 18, 27–44.

Droit-Volet, S. (1999b). L’estimation du temps: perspective développementale. L’AnnéePsychologique.

Droit, S., Pouthas, V., & Jacquet, A. Y. (1990). Temporal learning in 4 1/2- and 6-year-oldchildren: Role of instructions and prior knowledge. Journal of Experimental ChildPsychology, 50, 305–321.

Droit, S., Pouthas, V., & Jacquet, A. Y. (1991). Apprentissage temporel chez des enfantsâgés de 4 1/2 et 6 ans: Rôle d’une horloge externe. L’Année Psychologique, 91,347–364.

Fraisse, P. (1948). Etude comparée de la perception et de l’estimation de la durée chez lesenfants et chez les adultes. Enfance, 1, 199–211.

Fraisse, P. (1982). The adaptation of the child to time. In W.J. Friedman (Ed.), The devel-opmental psychology of time (pp. 113–140). New York: Academic Press.

Fraisse, P., & Orsini, F. (1958). Etude expérimentale des conduites temporelles: III Etudegénétique de l’estimation de la durée. L’Année Psychologique, 58, 1–6.

Friedman, W. J. (1990). Children’s representations to the daily activities. Child Develop-ment, 61, 1399–1412.

Hawkes, G. R., Warshan, R. W., & Ray, W. S. (1973). Response delay effects on durationjudgments for helicopter noise stimuli and vibration. Journal of Auditory Research,13, 26–30.

Karmiloff-Smith, A. (1992). Beyond modularity: A developmental perspective on cogni-tive science. Cambridge: MIT press.

Karmiloff-Smith, A. (1994). Precis of beyond modularity: A developmental perspectiveon cognitive science. Behavioral and Brain Sciences, 17, 693–745.

Levin, I. (1979). Interference of time-related and unrelated cues with duration comparisonof young children: Analysis of Piaget’s formulation of the relation of time andspeed. Child Development, 50, 469–477.

Levin, I. (1992). The development of the concept of time in children: An integrativemodel. In F. Macar, V. Pouthas, & W. J. Friedman (Eds.), Time, action and cogni-tion: Towards bridging the gap (pp. 13–33). Dordrecht: Kluwer Academic Publishers.

Logan, G. D. (1988). Toward an instance theory of automatization. Psychological Review,95, 4, 492–527.

Macar, F. (1988). Temporal regulation in children of 3- to 5-year-olds. Cahiers de Psy-chologie Cognitive, 8, 39–51.

Matsuda, F. (1991). A longitudinal study on a duration estimation: Effects of frequency ofintermittent stimuli. Japanese Psychological Research, 33, 48–52.

Matsuda, F., & Matsuda, M. (1983). A longitudinal study of learning process of durationestimation in young children. Japanese Psychological Research, 25, 3, 119–129.

Matsuda, F., Miyazaki, H., & Matsuda, M. (1983). Effects of style of response on learningof time estimation in young children and adults. Bulletin of the Faculty of HomeEconomics, 19, 119–129.

Michon, J. A. (1990). Implicit and explicit representations of time. In R. A. Block (Ed.),Cognitive models of psychological time (pp. 37–58). Hillsdale: Lawrence Erlbaum.


Montangero, J. (1977). La notion de durée chez l’enfant de 5 à 9 ans. Paris: Presses Uni-versitaires de France.

Montangero, J. (1979). La genèse des raisonnements et des concepts temporels. In P.Fraisse (Ed.), Du temps biologique au temps psychologique (pp. 175–215). Paris:Presses Universitaires de France.

Montangero, J. (1993). From the study of reasoning on time to the study of understandingthings in time. Psychological Belgica, 33, 185–195.

Mounoud, P. (1994). Le passage de représentations partielles à des représentations d’en-semble. Enfance, 1, 5–32.

Nelson, K. (1986). Event knowledge: Structure and function in development. Hillsdale,NJ: Erlbaum.

O’Connel, B., & Gerard, A. (1985). Scripts and scraps: The development of sequential un-derstanding. Child Development, 56, 671–681.

Ornstein, R. (1969). On the experience of time. Middlesex: Penguin.Piaget, J. (1946). Le développement de la notion de temps chez l’enfant. Paris: Presses

Universitaires de France.Piaget, J. (1974a). Réussir et comprendre. Paris: Presses Universitaires de France.Piaget, J. (1974b). La prise de conscience. Paris: Presses Universitaires de France.Pouthas, V., Droit, S., & Jacquet, A. Y. (1993). Temporal experiences and time knowl-

edge in infancy and early childhood. Time and Society, 2, 199–218.Pouthas, V., Droit, S., Jacquet, A. Y., & Wearden, J. (1990). Differentiation of response

duration in children of different ages: Developmental changes in relation betweenverbal and non verbal behavior? Journal of the Experimental Analysis of Behavior,53, 21–31.

Poynter, W. D. (1983). Duration judgment and the segmentation of experience. Memory &Cognition, 11, 77–82.

Richie, D. M., & Bickhard, M. H. (1988). The ability to perceive duration: Its relations tothe development of the logical concept of time. Developmental Psychology, 24,318–323.

Seymour, P. (1980). Internal representation of the months: An experimental analysis ofspatial forms. Psychological Research, 42, 255–273.

Siegler, S. R. (1978). Children’s thinking: What develops? Hillsdale NJ: Lawrence Erl-baum Associates.

Thomas, E. A. C., & Weaver, W. B. (1975). Cognitive processing and time perception.Perception and Psychophysics, 17, 363–367.

Wilkening, F., Levin, I., & Druyan, S. (1981). Integrating velocity, time and distance in-formation: A developmental study. Cognitive Psychology, 13, 231–247.

Zakay, D. (1989). Subjective and attentional resource allocation: An integrated model oftime estimation. In I. Levin & D. Zakay (Eds.), Time and Human Cognition (pp.365–397). Amsterdam: North-Holland.

Zakay, D., & Block, R. A. (1996). The role of attention in time estimation processes. InM. Pastor & J. Atieda (Eds.), Time, internal clocks and movement (pp. 143–164).Amsterdam: Elsevier Science.

Zakay, D., & Block, R. A. (1998). An attentional-gate model of prospective time estima-tion. In V. De Keyser, G. D’Ydevalle & A. Vandierendonck (Eds.), Time and theDynamic Control of Behavior (pp. 129– 143). Seattle: Hogrefe & Huber Publishers.


Zakay, D., & Fallach, R. (1984). Immediate and remote time estimation: A comparison.Acta Psychologica, 57, 69–81.

Zakay, D., Tsal, Y., Moses, M., & Shahar, I. (1994). The role of segmentation in prospectiveand retrospective time estimation processes. Memory & Cognition, 22, 344–351.

Zelazo, P. D., Frye, D., & Rapus, T. (1996). An age-related dissociation between knowingrules and using them. Cognitive Development, 11, 37–63.

Are Time and Action Dissociated in Young Children's Time Estimation?

Documents

Transcript of Are Time and Action Dissociated in Young Children's Time Estimation?