!!Development of Future Thinking in Young Children-semiopt-PrabhakarHudson, 2014

The development of futuchildrens ability to consto achieve future goals

Janani Prabhakar , Judith A. HuDepartment of Psychology, Rutgers, The State Univer

d goals. Results arefuture thinking inl reasoning ability,

hed by Else

http://dx.doi.org/10.1016/j.jecp.2014.02.004

Corresponding author.E-mail address: [email protected] (J. Prabhakar).

Journal of Experimental Child Psychology xxx (2014) xxxxxx

Contents lists available at ScienceDirect

Journal of Experimental ChildPsychology

journal homepage: www.elsevier .com/locate/ jecp0022-0965/ 2014 Published by Elsevier Inc.contingencies between the temporally orderediscussed in terms of the development ofconjunction with working memory, inferentiaand goal maintenance abilities.

2014 PublisPlease cite this article in press as: Prabhakar, J., & Hudson, J. A. The development of future thinking: Youdrens ability to construct event sequences to achieve future goals. Journal of Experimental Child Psychologyhttp://dx.doi.org/10.1016/j.jecp.2014.02.004vier Inc.Goal maintenanceto satisfy future goals. The results indicate that 4-year-olds wereable to successfully accomplish two temporally ordered goals evenwith high working memory demands and a complex goal structure,whereas 3-year-olds were able to accomplish two goals only whenthe working memory demands were low and the goal structure didnot involve additional demands from inferential reasoning anda r t i c l e i n f o

Article history:Available online xxxx

Keywords:Future-oriented behaviorSemantic future thinkingMental time travelWorking memoryPlanningre thinking: Youngtruct event sequences

dsonsity of New Jersey, New Brunswick, Piscataway, NJ 08854, USA

a b s t r a c t

Previous studies suggest that the ability to think about and act onthe future emerges between 3 and 5 years of age. However, it isunclear what underlying processes change during the developmentof early future-oriented behavior. We report three experimentsthat tested the emergence of future thinking ability throughchildrens ability to explicitly maintain future goals and constructfuture scenarios. Our main objectives were to examine the effectsof goal structure and the effects of working memory demands onchildrens ability to construct future scenarios and make choicesng chil-(2014),

AlexandraEvideniere

2 J. Prabhakar, J.A. Hudson / Journal of Experimental Child Psychology xxx (2014) xxxxxxIntroduction

Humans are able to envision their past and think about possible future events. This ability, calledmental time travel, allows individuals to construct details that occurred in a past event and will occurin a future event (Suddendorf & Corballis, 2007; Tulving, 2002). Furthermore, with mental time travel,individuals can use details from past events to generate possible future actions. Future-directedthought is essential for complex cognitive activities, including planning for the future, accomplishinggoals, and remembering to complete intended actions, that is, prospective memory. The study of theroots of future thinking behavior has gained much attention during the last decade primarily througha focus on childrens ability to make choices in the present to satisfy future needs (Atance & ONeill,2005b; Lemmon & Moore, 2007; McColgan & McCormack, 2008; Metcalf & Atance, 2011; Russell,Alexis, & Clayton, 2010; Suddendorf, Nielson, & von Gehlen, 2011). This recent focus on childrenschoices for the future has allowed researchers to focus on the emergence of future-directed thoughtduring preschool years as a basis for more complex future-oriented behavior.

Raby and Clayton (2009) distinguished distinct categories that fall under the umbrella of future-oriented behavior. They suggested that prospective thinking reects a declarative form of future-ori-ented thought and can be subdivided into prospective, semantic, and episodic future thinking. In theirdescription, semantic future thinking is similar to semantic memory in that it does not necessarilyinvolve projecting oneself into the future. That is, with semantic future thinking, individuals can imag-ine a future event using script-based or general event knowledge. In contrast, with episodic futurethinking, individuals are part of the imagined future event and where, when and what aspects areincorporated into the future event. However, when individuals imagine the future, they likely drawon both semantic and episodic future-oriented thought processes. For example, when imagining afuture birthday party, individuals may incorporate elements that are likely to play a role in the partysuch as a birthday cake, and also plan for novel circumstances of the specic future event such as thebest route to reach the party.

In terms of development, past research suggests that although semantic memory and episodicmemory processes may emerge together, children are able to use semantic information to plan forfuture events earlier than episodic information (Hudson, 1986; Hudson & Fivush, 1991; Hudson,Shapiro, & Sosa, 1995; Kreitler & Kreitler, 1987; Nelson & Gruendel, 1981, 1986; Slackman, Hudson,& Fivush, 1986). That is, although young children are able to use general event knowledge to describefuture situations, they nd it difcult to use specic event details to plan for novel future events(Hudson et al., 1995; Kreitler & Kreitler, 1987). In addition, prior research suggests that episodic futurethinking is a precursor for more complex future-oriented behavior, specically prospective memory(Nigro, Brandimonte, Cicogna, & Cosenza, in press). Prospective memory requires individuals toemploy past- and future-directed thinking. Individuals must remember intended actions and also planon how to accomplish them (Einstein & McDaniel, 1990; Graf & Uttl, 2001), taxing working memoryand cognitive exibility.

The primary goal of this study was to extend knowledge regarding early future-oriented behavior inpreschool-age children in several importantways. First, we sought to determine childrens ability to usesemantic knowledge to accomplish a set of future goals. Although 3-year-old children are able to usescript-based knowledge to describe future events (Hudson & Fivush, 1991; Hudson et al., 1995), thereis little information as towhether children at this age are able to use semantic knowledge to accomplishspecic future goals. Second, this studywould allow us to determine the nature of early future-orientedbehavior. Prior research suggests that future-oriented behavior emerges between 3 and 5 years of age(see Hudson, Mayhew, & Prabhakar, 2011, for a review). These studies have largely focused on youngchildrens ability to place events accurately on a temporal scale, provide verbal reports of personal pastand future events, andmake choices that satisfy personal future goals (Atance &ONeill, 2005a; Busby &Suddendorf, 2005; Grant & Suddendorf, 2010; Hudson & Mayhew, 2011; McCormack & Hanley, 2011).Many of these studies have found rudimentary past and future thinking ability in 3-year-olds and adevelopmental progression that culminates in a much more complex ability between 4 and 5 years ofage. For example, 3-year-olds are able to converse about the real past and future using temporal termssuch as before and after or did and will (Blewitt, 1982; Busby & Suddendorf, 2005; Grant & Suddendorf,Please cite this article in press as: Prabhakar, J., & Hudson, J. A. The development of future thinking: Young chil-drens ability to construct event sequences to achieve future goals. Journal of Experimental Child Psychology (2014),http://dx.doi.org/10.1016/j.jecp.2014.02.004

AlexandraEvideniere

AlexandraEvideniere

AlexandraEvideniere

AlexandraEvideniere

AlexandraEvideniere

AlexandraEvideniere

AlexandraEvidenieredef

2010, 2011). However, they are only able to describe events that fall ona specicday (yesterdayor tomor-row) at the age of 4 (Busby&Suddendorf, 2005;Grant&Suddendorf, 2010, 2011). Furthermore, although3-year-olds are able to use modal auxiliaries such as going to, want to, should, and could (Harner, 1975,1982), the ability to plan actions or make accurate choices to satisfy future needs does not emerge until4 or, in some cases, 5 years of age (Atance & Jackson, 2009; Atance & Meltzoff, 2005; Atance & ONeill,2005a; Busby & Suddendorf, 2005; McColgan & McCormack, 2008; Moore & Lemmon, 2001; Russellet al., 2010; Suddendorf et al., 2011).

The majority of past research on future thinking has asked children to make choices for specicfuture events. For example, Atance and ONeill (2005b) asked 3-year-olds to select three objects totake with them on a trip and then asked the children to explain why they chose each item. Theresearchers found that children as young as 3 years were able to refer to the future in their explana-tions and, furthermore, could account for uncertainties that were present when planning for the future(e.g., bringing a bandage in case someone gets hurt). Because children were able to incorporate uncer-tainty into their explanations (e.g., accounting for the possibility of getting hurt), these ndings sug-gest that 3-year-olds have the ability to think about and refer, linguistically, to the future in thechoices they make. Furthermore, childrens choices may have been based on semantic future thinkingprocesses (e.g., choosing a bandage because it is typically used when someone is hurt). In a similarstudy, Atance and Meltzoff (2005) asked children to bring one of three items to a future location(e.g., the mountains, the desert). The correct item in this study served a functional purpose (e.g., takingsunglasses to the sunny desert). Unlike Atance and ONeill (2005b), this study found that younger chil-dren had difculty in explaining their choices. The authors suggested that although younger childrenmay have language difculties that impede their explanatory behavior, the larger issue is the tendencyto select objects based on associations. Therefore, in Atance and Meltzoff (2005) study, althoughyounger children were able to select the functional tool, they less often used causal language (e.g.,I chose the sunglasses because the sunny desert will cause sun in my eyes). Instead, they tendedto use language that reected associations (e.g., I chose the sunglasses because they can be wornin a sunny desert).

These types of responses do not necessarily indicate a lack of future thinking ability; rather, theyindicate a tendency to make future-directed choices based on semantic associations. Atance and col-leagues described childrens ability to arrive at correct semantic associations, but it is not knownwhether children can use semantic knowledge to plan for and accomplish future goals. To ensure thatchildren address questions with the future in mind, other researchers have created novel scenarios inorder to study childrens ability to extract information from specic past events to make future-direc-ted choices. Suddendorf and colleagues (2011) rst showed children an apparatus that yielded asticker when using a particular tool and then, either immediately or after a delay, asked children toselect one tool from a group of tools. Importantly, children were not told for what purpose theyneeded to choose a tool and were expected to infer the intentions based on their previous experience(e.g., to get a sticker from the apparatus). The researchers found that 4-year-olds chose the correct toolsignicantly more than by chance during both delay conditions, whereas 3-year-olds were abovechance only when asked immediately. These results suggest that 3-year-olds are capable of makingchoices for novel future scenarios. However, younger children could not maintain this ability withthe increased recall demands of the delay condition. One explanation is that because children werenot explicitly given the goal (to obtain a sticker), they were unable to retrieve the relevant knowledgefrom the past to accomplish the goal.

Russell and colleagues (2010) gave children an explicit goal and asked them to make a future-direc-ted choice. To make the choice, children needed to envision a past experience with a game (a blowfootball game) from a different visual perspective. They found that 4-year-olds were more accurate atchoosing the correct objects than 3-year-olds and, furthermore, that 4-year-olds were more accuratewhen choosing for someone else who would play tomorrow than when choosing for themselves toplay tomorrow. The authors suggested that children had difculty in imagining being in a differentspatial orientation, but doing so from another individuals perspective posed fewer spatial demands.These results indicate that spatial reorientation requires additional cognitive processing when simu-lating a future event. However, it is not clear whether 3-year-olds poor performance was due to dif-

J. Prabhakar, J.A. Hudson / Journal of Experimental Child Psychology xxx (2014) xxxxxx 3culties in future simulation or in spatial perspective taking.

Please cite this article in press as: Prabhakar, J., & Hudson, J. A. The development of future thinking: Young chil-drens ability to construct event sequences to achieve future goals. Journal of Experimental Child Psychology (2014),http://dx.doi.org/10.1016/j.jecp.2014.02.004

AlexandraEvideniere

AlexandraEvideniere

AlexandraEvideniere

AlexandraEvideniere

AlexandraEvideniere

AlexandraEvideniere

AlexandraEvideniere

AlexandraEvideniere

AlexandraEvideniere

AlexandraEvideniere

AlexandraEvideniere

AlexandraEvideniere

AlexandraEvideniere

Taken together, these studies show that 3-year-olds are able to use semantic associations to makechoices for a hypothetical future event (Atance & Meltzoff, 2005; Atance & ONeill, 2005b) and are ableto make correct choices for future needs in an environment they had just experienced (Suddendorfet al., 2011). However, these studies show that 3-year-olds have limited capacity to make future-directed choices and accomplish future goals when tasks also include additional temporal and spatialdemands. Therefore, studies that measure early future-oriented behavior should minimize thesedemands and focus on childrens ability to maintain and accomplish future goals. Many researcherscontend that knowledge about the future centers on personal goals (DArgembeau & Mathy, 2011),and future thinking involves intentionally acting on those goals (Szpunar, 2010). This suggests thatfuture thinking is an adaptive process, and our inclination to think about the future is derived fromthe desire to accomplish future goals. Therefore, studies of basic future thinking should examinethe maintenance of future scenarios based on clear indications of future goals.

In our study, we tested the emergence of future thinking abilities through childrens ability tomaintain future goals during the developmental shift between 3 and 4 years of age. Our main objec-tives were to examine the effects of goal structure and the effects of task demands on childrens futurethinking. We did not specify an exact future time to ensure that young childrens performance wouldnot be deated by a lack of understanding of temporal terminology because past research has indi-cated that 3-year-olds have a more limited understanding of future terms, such as tomorrow and yes-terday, than 4-year-olds. Because children as young as 3 years are able to construct future narrativesusing temporally nonspecic future-directed language, such as going and will, we included this type oflanguage to refer to the future in our instructions (Grant & Suddendorf, 2011).

Our study required children to use semantic future thinking rather than episodic future thinkingbecause past research indicates that although semantic and episodic memory processes may emergetogether, semantic future thinking may emerge earlier (Hudson, 1986; Hudson & Fivush, 1991;Hudson et al., 1995; Kreitler & Kreitler, 1987; Nelson & Gruendel, 1981, 1986; Slackman et al.,1986). In addition, past research has indicated that children have difculty in accomplishing taskswith high episodic demands (Russell et al., 2010; Suddendorf et al., 2011). Therefore, our goal wasto track childrens future-directed choices using past knowledge through their ability to accomplishexplicit future goals. In Experiment 1, children were asked to complete a step-by-step set of actionsrequiring completion of a subgoal and then a nal primary goal in the correct temporal order. By giv-ing children two goals to accomplish, we tested their ability to hold multiple goals in mind withoutplacing demands on children to generate goals on their own, as in Suddendorf and colleagues(2011) study. Because the task was in view of the children during the decision-making process, wewere also able to reduce spatial demands, as seen in Russell and colleagues (2010) study.

In our study, 3- and 4-year-old children were asked to carry out two goals by choosing the correctlocations to visit in a pretend neighborhood. These goals were linked by a temporal structure wherethe rst goal (the subgoal) must be accomplished before the second goal (the nal goal). We speci-cally varied the locations in the environment. The neighborhood had distinct locations, and childrenwere asked to choose the location that satised the subgoal and the nal goal. We kept spatial andmemory demands low and, instead, varied only the number of features in the environment. Wehypothesized that the need to consider more elements would require greater working memoryresources to construct future event scenarios and meet future goals. We predicted that when workingmemory demands are lower, younger children would benet from having fewer features to constructin the future scenario and would be better able to accomplish both goals. We further predicted thatwhen working memory demands are higher, only 4-year-olds would succeed in accomplishing bothgoals. Because we minimized the role of other cognitive processes (i.e., spatial perspective), we wereable to strip the future thinking process solely to childrens ability to choose features in the environ-ment that would satisfy explicit future goals.

We adopted language constructions similar to those used in past studies (Grant & Suddendorf,2011; Russell et al., 2010); that is, our instructions used auxiliary verbs in their future tense (e.g.,will, going, go). Feature demands were manipulated by changing the number of locations inthe pretend neighborhood (houses and stores). In the high demand condition the neighborhoodincluded six buildings, whereas in the low demand condition it included four buildings. Children were

4 J. Prabhakar, J.A. Hudson / Journal of Experimental Child Psychology xxx (2014) xxxxxxalso divided into different perspective groups (self vs. other) as in Russell and colleagues (2010) study.


AlexandraEvideniere

AlexandraEvideniere

AlexandraEvideniere

AlexandraEvideniere

AlexandraEvideniere

AlexandraEvideniere

AlexandraEvideniere

AlexandraEvideniere

AlexandraEvideniere

AlexandraEvideniere

AlexandraEvideniere

Children in the self condition were asked how they would accomplish two future goals, and children inthe other condition were asked how another child would accomplish the same two goals. This manip-ulation provided a basis for comparison between making future decisions for ones own future and foranother individuals future. We predicted that children would be better at making decisions for them-selves because needing to project the future of another individual may cause greater demands.

Experiment 1

Method

ParticipantsParticipants were 124 children: 60 3-year-olds (M = 42 months, SD = 3.77; 25 boys and 35 girls)

and 64 4-year-olds (M = 54 months, SD = 3.34; 34 boys and 30 girls). Participants were predominantlyWhite native English speakers from middle-class backgrounds. They were recruited from preschoolsnear New Brunswick, New Jersey, on the U.S. east coast. Parents of the children in these schools pro-vided consent for their children to participate, and children were given stickers for participation.

J. Prabhakar, J.A. Hudson / Journal of Experimental Child Psychology xxx (2014) xxxxxx 5MaterialsMaterials included a 0.91 1.22-m model of a neighborhood (see Figs. 1 and 2). Each model con-

tained different numbers of miniature houses and stores around a black road. The number of housesand stores in the neighborhood varied depending on feature demand conditions. Each child was givenone of two dolls; boys were shown a boy doll, and girls were shown a girl doll. These dolls were usedto move from one location to another in the neighborhood based on childrens choices. A video camerawas also used to record each session.

Design and procedureChildren in all conditions were introduced to the model neighborhood with its toy houses and

stores. Buildings were labeled by their color and their name (e.g., The red house is Marys house).All locations were labeled in random order for participants without any one being identied as keyto the future decision. The neighborhood was designed such that all houses were distinct from allstores in terms of shape, and all locations were distinct from one another by color. Half of the childrenwere randomly assigned to the low demand condition (see Fig. 1), and the remaining children wereassigned to the high demand condition (see Fig. 2). In both conditions, there was one target house(the nal goal location) and one target store (the subgoal location). Therefore, there were two dis-tracter locations in the low demand condition and four in the high demand condition. In the low

Fig. 1. Image of the three-dimensional neighborhood in low demand. This image reects the low demand condition, with twotarget locations and two distracter locations. Children were given either the boy doll or the girl doll to move around the

neighborhood.


AlexandraEvideniere

AlexandraEvideniere

AlexandraEvideniere

AlexandraEvideniere

AlexandraEvideniere

6 J. Prabhakar, J.A. Hudson / Journal of Experimental Child Psychology xxx (2014) xxxxxxdemand condition, participants were shown three houses and one store. The starting and nal goallocations were located on opposite sides and were counterbalanced by participant. The toy store(the subgoal location) and the remaining house were on opposites sides and were also counterbal-anced by participant. In the high demand condition, participants were shown four houses and twostores. The two stores were on opposite sides and counterbalanced by participant. The starting andnal goal locations were also on opposite sides, with their individual locations counterbalanced withthe remaining two houses by participant. All participants started by placing a doll in front of the pur-ple house (starting location) and were given two goals in sequential order (a subgoal and a nal goal).Participants were asked to move the doll to the location that satised each goal.

Within each demand condition, half of the children were asked to make future choices for them-selves (self condition) and the remaining children were asked to make future choices for another indi-vidual (other condition). In the self condition children were asked to pretend that the doll wasthemselves, and children in the other condition were asked to pretend that the doll was another child(Charlie or Jill).

Once the experimenter had indicated the name and color of each location, children were givenpractice trials to test their memory for the locations in the neighborhood. Children were rst asked

Fig. 2. Image of the three-dimensional neighborhood in high demand. This image reects the high demand condition, with twotarget locations and four distracter locations. Children were given either the boy doll or the girl doll to move around theneighborhood.to identify each house and store (e.g., Can you showme where Bobs house is?). They were correctedif they incorrectly answered this question. Children were also asked to walk the doll from location tolocation (e.g., in the self condition: Show me how you walk to Bobs house; in the other condition:Showme how Charlie walks to Bobs house). The order of identication was randomized. By the sec-ond practice trial, all participants had correctly identied each house or store. Children were then pro-vided with the two goals. The script for the goals was generated to cohere with preschoolerscomprehension and use of future-directed language. Past research has shown that by 3 years of agechildren are able to refer to the future and construct sentences using modal auxiliaries such as willand going (Grant & Suddendorf, 2010; Harner, 1982; Hayne, Gross, McNamee, Fitzgibbon, & Tustin,2011). Therefore, we incorporated simple modal auxiliaries to ensure that language ability would notaffect understanding of future-directed action. All children were told, Today is Marys birthday, andMary is going to have a birthday party at her house. They were then given the two goals in temporalsequencerst the subgoal and then the nal goal. In this way, the temporal structure was providedfor children, allowing us to analyze the completion of each goal separately. Children in the self condi-tion were rst told, You want to give Mary a birthday present. Where will you go rst?.

To be credited with a correct response, children needed to both point to and verbally identify thelocation for the subgoal (the toy store). If children pointed to an incorrect location, they were imme-diately told the correct location (the toy store) and were asked to walk the doll to the toy store ratherthan their incorrect selection. As a result, before the nal goal prompt was given, all children had a


AlexandraEvideniere

AlexandraEvideniere

AlexandraEvideniere

AlexandraEvideniere

AlexandraEvideniere

AlexandraEvideniere

AlexandraEvideniere

AlexandraEvideniere

AlexandraEvideniere

present from the toy store. With the doll standing by the toy store, children were then given the nalgoal prompt: You now have a present for the birthday party. Where will you go next? Again, to becredited with a correct response, children needed to both point to and verbally identify the location forthe nal goal (Marys house). All children were given the subgoal and nal goal prompts at the samelocations. The starting location was the same for all children, and all children were directed to the toystore to get a present before being given the nal goal. Children in the other condition were given thesame scenario and prompts but were told to make their choice for another individual, either Charlie or

A 2 (Age Group: 3- vs. 4-year-olds) 2 (Feature Demands: high vs. low) 2 (Goal: subgoal vs. nal

performed to compare performance on each question across conditions at both ages. As shown in

J. Prabhakar, J.A. Hudson / Journal of Experimental Child Psychology xxx (2014) xxxxxx 7Fig. 3, all children were more accurate in selecting the correct nal goal when the demand was lowthan when the demand was high, t(120) = 2.07, p < .05, d = 0.38. As shown in Fig. 4, 3-year-olds weremore accurate in selecting the correct location for the nal goal than for the subgoal, t(111) = 5.13,p < .001, d = 0.94, but this difference was not signicant for 4-year-olds, t(126) = 0, p = 1.

Table 1Means for independent variables by goal.

Subgoal Final goal

3-year-olds, low demand .17 .673-year-olds, high demand .20 .534-year-olds, low demand .63 .784-year-olds, high demand .72 .56goal) mixed factorial ANOVA was performed on the number of correct location choices (see Table 1 formeans). Type of goal was a within-participant variable, and age, feature demands, and perspectivewere between-participant variables. The ANOVA yielded signicant main effects of age, F(1,120) = 20.04, p < .001, g2G = .08, and goal, F(1, 120) = 14.20, p < .001, g2G = .04, as well as signicant inter-actions between demand and goal, F(1, 120) = 5.11, p = .01, g2G = .02, and between age and goal, F(1,120) = 15.14, p < .001, g2G = .05. Despite the fact that ANOVAs are robust to violations of normality, weran a logistic regression on these data because our dependent measure was binomial. A parallel logisticregression with age, goal, and demand as predictors found similar results to the ANOVA (Table 2).

The 4-year-olds were signicantly more accurate in selecting the correct location for both goals(M = .67, SD = .47) than the 3-year-olds (M = .39, SD = .49), and both age groups were more accuratein selecting the correct nal goal (M = .64, SD = .48) than in selecting the subgoal (M = .44, SD = .50).To further examine the interactions between demand and goal, and between age and goal, t tests wereJill. None of the instructions for any of the conditions asked children explicitly to imagine scenarios.Rather, in all conditions, children were given explicit goals (i.e., You want to give Mary a birthdaypresent or Jill wants to give Mary a birthday present) and were asked what actions they or Jillwould take. For every correct answer, children were given 1 point.

Results

In this study, we were interested in the effect of age (3- vs. 4-year-olds), feature demands (low vs.high), goal (subgoal vs. nal goal), and perspective (self vs. other) on childrens ability to choose thecorrect location to satisfy future goals. An analysis of variance (ANOVA) is reported to determinethe effect of each of the independent variables on overall accuracy. Even though accuracy was adichotomous variable, an ANOVA was conducted instead of a nonparametric test because it is ableto yield interactive effects and is robust to violations of assumptions (Schmider, Ziegler, Danay,Beyer, & Bhner, 2010). Nonetheless, we also included a logistic regression to substantiate the resultsof the ANOVA. Additional analyses were conducted to understand childrens pattern of choices for thesubgoal and nal goal questions. Preliminary analyses found no effects of gender or perspective, andso analyses were collapsed across these variables.

Overall accuracyPlease cite this article in press as: Prabhakar, J., & Hudson, J. A. The development of future thinking: Young chil-drens ability to construct event sequences to achieve future goals. Journal of Experimental Child Psychology (2014),http://dx.doi.org/10.1016/j.jecp.2014.02.004

AlexandraEvideniere

AlexandraEvideniere

AlexandraEvideniere

AlexandraEvideniere

AlexandraEvideniere

AlexandraEvideniere

AlexandraEvideniere

AlexandraEvideniere

AlexandraEvideniere

AlexandraEvideniere

AlexandraEvideniere

AlexandraEvideniere

AlexandraEvideniere

AlexandraEvideniere

Table 2Logistic regression with age, demand, and goal acting as predictors of accuracy.

0.8

8 J. Prabhakar, J.A. Hudson / Journal of Experimental Child Psychology xxx (2014) xxxxxx0.2

0.4

0.6b SE b z p

Age 4.05 0.84 3.13

Table 3Percentages of children selecting possible options for the subgoal location.

Age (years) Demand Incorrect Incorrect CorrectFinal goal location Distracter location Toy store

3 (n = 29) High 14 66 213 (n = 30) Low 53* 30 174 (n = 32) High 6 22 72*

4 (n = 28) Low 18 11 71*

Note. Chance levels differ by demand condition (high: 1/6 for nal goal location and toy store, 4/6 for distracter location; low: 1/4 for nal goal location and toy store, 2/4 for distracter location).* Different from chance by an exact binomial test (p < .05).

J. Prabhakar, J.A. Hudson / Journal of Experimental Child Psychology xxx (2014) xxxxxx 9signicantly more than by chance (low demand: chance = 1/4; high demand: chance = 1/6; exact bino-mial tests, p < .01).

Whereas 4-year-olds successfully chose the correct location for the subgoal, 3-year-olds were lesssuccessful and more affected by feature demands. Specically, when the demand was low, theyseemed to ignore the subgoal and selected the location that satised the nal goal. However, in thehigh demand condition, 3-year-olds selected locations randomly.

Responses to nal goal questionTo examine whether responses to the nal goal question were dependent on performance on the

subgoal location question, two Spearmans rank correlation tests were performed on the correctresponse scores for the subgoal and nal goal questions. No signicant relationship was foundbetween accuracy on the subgoal and nal goal questions in the high demand condition (r = .14,p = .29) or low demand condition (r = .21, p = .10).

Analyses were performed on responses to the nal goal question to verify that children chose thecorrect location at levels greater than chance (Table 4). Both younger and older children chose the cor-rect nal goal location signicantly more than by chance in both the high demand condition(chance = 1/6; exact binomial test, p < .001) and low demand condition (chance = 1/4; exact binomialtest, p < .001). In addition, 3-year-olds in the low demand condition who incorrectly selected Maryshouse in response to the subgoal question (n = 16) were above chance in selecting Marys house forthe nal goal (n = 12) (chance = 1/4; exact binomial test, p < .001).

Discussion

We predicted that 3-year-olds would successfully complete future goals when working memorydemands were minimal but that 4-year-olds should be successful in all conditions. The results par-tially corroborated our predictions. Overall, all children were more accurate when the featuredemands were low, and 4-year-olds were more successful in completing both goals than 3-year-olds.In contrast, 3-year-olds were more successful in completing the nal goal than in completing theTable 4Percentages of children selecting possible options for the nal goal location.

Age Demand Incorrect Incorrect CorrectStarting location Distracter location Marys house

3 years High 15 26 59*

3 years Low 20 13 67*

4 years High 16 28 56*

4 years Low 14 86*

Note. Chance levels differ by demand condition. When demand is high, chance for individual locations (i.e., Marys house, toystore) is 1/6 and for distracter locations is 4/6. When demand is low, chance for individual locations is 1/4 and for distracterlocations is 2/4.* Different from chance by an exact binomial test (p < .05).


AlexandraEvideniere

in thinking about the subgoal. There are several possible explanations. One explanation is that an

immature inhibition mechanism may have made it difcult for younger children to suppress theprimary goal in order to complete the subgoal. That is, they might not have been able to inhibit theirpreference to go to the birthday party at Marys house due to an immature inhibition mechanism.However, it is unclear why children might have a preference for a primary goal over a subgoal. Onepossibility is that 4-year-olds have better access to general event knowledge and are more aware thatone should buy a birthday present before attending a birthday party. Although this is possible, pastresearch indicates that 3- and 4-year-olds are equally capable of using general event knowledge tocreate plans such as going to a grocery store or to the beach (Hudson et al., 1995). Therefore, it isunclear whether younger children in Experiment 1 were unable to access their general eventknowledge about birthdays to complete the subgoal. In Experiment 2, we asked children, explicitly,where they would go to get a birthday present to address the possibility that younger children mightnot have access to such general event knowledge.

Another explanation for younger childrens behavior in Experiment 1 could be that 3-year-oldschose Marys house more often for the subgoal because they believed that they needed to go to Maryshouse to give her something but not necessarily a present. However, the task instructions made it clearto children that they wanted to get Mary a birthday present and, as a result, the action must be to getthe present rst. What this suggests is that 3-year-olds might not have inferred a goal-directed mentalstate automatically. That is, when they were told that they wanted to get Mary a birthday present, 3-year-olds might not have inferred that the mental state, want, suggests a goal to be accomplished.Experiment 2 was also conducted to determine whether younger children are able to automaticallyinfer the goal from the instructions. To fulll the subgoal, children needed to make an inference asto the location that would allow them to buy a present, whereas the nal goal required memory fora single location introduced previously. Making an inference may be more difcult than rememberinga location, which could explain why 3-year-olds performance was better on the nal goal questionthan on the subgoal question. To explore these possibilities, Experiment 2 tested childrens abilityto make the explicit inference that the toy store would provide presents in a task that was free of addi-tional future decisions.

In addition to the possibility that younger children were unable to make a goal-directed inference,in Experiment 1, younger childrens difculty in accomplishing the subgoal could also be attributed tothe contingent relationship between the subgoal and the nal goal. The two goals were inherentlylinked to one another in a temporally ordered structure such that the subgoal was required to becompleted before the nal goal. It could be the case that younger children found it difcult to planactions in such a contingent manner. In Experiment 3, we tested childrens ability to accomplishtwo explicit goals that did not require an inference or contain any contingencies.

Because 3-year-olds tended to skip to the nal goal in the low demand condition in Experiment 1,in Experiments 2 and 3 the 3-year-olds were tested only in the low demand condition. Furthermore,because 4-year-olds were accurate in low demand and faced greater difculties when the demand wasincreased, in Experiments 2 and 3 the 4-year-olds were tested only in the high demand condition.Finally, because we found no differences between the self and other conditions in Experiment 1, onlythe self condition was used in Experiments 2 and 3.

Experiment 2

In Experiment 2, we tested childrens ability to make an inference regarding which location to visitin order to buy a birthday present. If children are able to accomplish this task, we could conclude thatyounger children can draw the correct explicit inference and their poor performance on the subgoalsubgoal and tended to skip directly to accomplish the nal goal. These results indicate that althoughyounger children were affected by higher feature demands, the type of goal was also a limiting factor.Specically, they had difculty in completing the embedded subgoal.

We tested childrens ability to accomplish future goals using semantic future thinking (Raby &Clayton, 2009). The 3-year-olds were able to satisfy a single nal (primary) goal but had difculty

10 J. Prabhakar, J.A. Hudson / Journal of Experimental Child Psychology xxx (2014) xxxxxxquestion in Experiment 1 was not due to lack of inferential ability.


AlexandraEvideniere

AlexandraEvideniere

If children correctly selected the toy store by pointing to it and verbally identifying it, they were given

1 point.

Results and discussion

Both 4-year-olds (M = .87) and 3-year-olds (M = .80) chose the toy store signicantly above chance(p < .001 by an exact binomial test). Thus, both age groups were able to make the correct inferencewhen they were explicitly asked where they would go to buy a birthday present in the pretend neigh-borhood. These results suggest that 3- and 4-year-olds are equally capable of drawing on generalevent knowledge (i.e., script knowledge) to address future goals. In addition, these results suggest thatin Experiment 1 the 3-year-olds difculty with the subgoal question was not due to their inability todraw the appropriate inference (that they needed to go to the toy store to buy a birthday present);rather, they may have had difculty in accomplishing two contingent, temporally ordered future goals(getting the birthday present rst so that they could take the present to the birthday party later). Totest childrens ability to consider two future goals, children in Experiment 3 were asked to accomplishtwo future goals that were mutually exclusive, were temporally ordered, and did not involve anyinferences.

Experiment 3

Method

ParticipantsParticipants were 16 3-year-olds (M = 42 months, SD = 2.93; 10 girls and 6 boys) and 16 4-year-olds

(M = 54 months, SD = 7.09; 9 girls and 7 boys). Participants were predominantly White native Englishspeakers from middle-class backgrounds. They were recruited from preschools near New Brunswickas in Experiment 1. Parents of the children in these schools provided consent for their children toparticipate, and children were given stickers for participation.

Materials and procedureThe same neighborhood and dolls from Experiment 1 were used. The 3-year-olds were tested with

the low demand layout, and the 4-year-olds were tested with the high demand layout. Children weregiven the same practice trials as in Experiment 1. After successful completion of the practice trials,children were given two goals: (a) to go to a birthday party and (b) to go to a sleepover. They weretold, Today you are going to a birthday party at Marys house, and then you are going to a sleepoverMethod

ParticipantsParticipants were 15 3-year-olds (M = 41 months, SD = 2.70; 8 girls and 7 boys) and 15 4-year-olds

(M = 53 months, SD = 3.44; 7 girls and 8 boys). Participants were predominantly White native Englishspeakers from middle-class backgrounds. They were recruited from preschools near New Brunswickas in Experiment 1. Parents of the children in these schools provided consent for their children toparticipate, and children were given stickers for participation.

Materials and procedure

The same neighborhood and dolls from Experiment 1 were used. The 3-year-olds were tested withthe low demand layout, and the 4-year-olds were tested with the high demand layout. Children weregiven the same practice trials as in Experiment 1. After successful completion of the practice trials,children were told, Today is Marys birthday, and you want to get Mary a birthday present. Wherewill you go to get a birthday present for Mary? Although this question was phrased in the futuretense, the question itself made it clear that the end goal was to get a birthday present for Mary today.

J. Prabhakar, J.A. Hudson / Journal of Experimental Child Psychology xxx (2014) xxxxxx 11at Sams house. Where will you go rst? Children were corrected if they were wrong. After they


AlexandraEvideniere

AlexandraEvideniere

AlexandraEvideniere

AlexandraComentariu pt. text

AlexandraEvideniere

AlexandraEvideniere

AlexandraEvideniere

AlexandraEvideniere

AlexandraEvideniere

AlexandraEvideniere

Table 5 shows younger and older childrens performance for both goal questions in Experiments 1Because past research has shown evidence of differential future thinking abilities between 3- and4-year-old children (Atance & ONeill, 2005a, 2005b; Grant & Suddendorf, 2010, 2011; Russell et al.,2010; Suddendorf et al., 2011), the goal of this investigation was to measure early future-orientedbehavior in young children. We examined preschoolers ability to envision the future in order to com-plete goals and subgoals under varying task demands. Specically, we hypothesized that 3-year-oldsability to plan for future event goals would be affected by working memory demands of the futureenvironment. Our results indicate that, as expected, 4-year-olds were able to accomplish twoand 3. All children were able to accomplish both goals when neither goal required an additional infer-ence; both 3- and 4-year-olds were above chance when selecting the rst and second goals in low andhigh demand, respectively (p < .001 by an exact binomial test). Four-year-olds performance in Exper-iment 3 was not signicantly different from 4-year-olds performance in Experiment 1 on the subgoalquestion, t(25) = 0.35, p = .73, and the nal goal question, t(33) = 1.71, p = .10. This indicates that olderchildren were able to accomplish both goals, with and without inferences, in high and low demandconditions. On the other hand, 3-year-olds were able to accomplish both goals only when neitherinvolved an inference and when the demand was low (as in Experiment 3).

An interesting point to note is the nonsignicant difference between 4-year-olds performance onthe nal goal in Experiment 3 versus that in Experiment 1. Four-year-olds were more accurate in com-pleting the nal goal when given noncontingent goals (Experiment 3) than when given contingentgoals (Experiment 1). Further research is required to determine whether there is a difference in thecognitive mechanisms that underlie future-directed processes toward accomplishing contingent ver-sus noncontingent goals.

General discussionmoved the doll to the correct house, they were told to pretend that the birthday party was over andthey were asked, Where will you go next?

The location of each of these events (Marys house or Sams house) was counterbalanced betweenparticipants. However, the temporal order was always the same (i.e., birthday party and then sleepov-er). If children selected the correct location by pointing to it and verbally identifying it, they weregiven 1 point.


Table 5Childrens accuracy in selecting correct location for each goal in Experiments 1 and 3.

Experiment Age (years) Demand Goal 1 accuracy Goal 2 accuracy

1 3 Low .17 .671 4 High .72 .563 3 Low .80 .943 4 High .67 .80

12 J. Prabhakar, J.A. Hudson / Journal of Experimental Child Psychology xxx (2014) xxxxxxembedded future goals even when working memory demands were increased. On the other hand,3-year-olds ability to plan for the future event goals was inuenced by the cognitive complexitiesof the task, including the working memory demands, the temporal nature of the event structure,and the inferential reasoning demands. Although they were not successful in accomplishing two tem-porally ordered goal actions that required an inference even when working memory demands wereminimized, 3-year-olds were able to plan for two future goals when the goals were explicit and notcontingent.

Three-year-olds difculty in mentally representing multiple goals does not necessarily mean thatchildren at this age are incapable of semantic future thinking; rather, it could mean that youngerchildrens ability to engage in future-oriented behavior is very limited. The results of the current study


AlexandraEvideniere

AlexandraEvideniere

AlexandraEvideniere

AlexandraEvideniere

AlexandraEvideniere

AlexandraEvideniere

AlexandraEvideniere

AlexandraEvideniere

suggest that 3-year-olds have a nascent, but tenuous ability to mentally maintain and enact futuregoals. Notably, they were successful only when working memory demands were minimized by includ-ing fewer distracter items in the future goal environment. However, reducing distracter items was notsufcient to produce accurate performance if they were also required to draw an implicit inferenceand appreciate the contingent relationship between the two goals. It was not the case that 3-year-oldswere incapable of drawing inferences when the goal was made explicit (Experiment 2), but the cog-nitive demands of inferential reasoning and planning for both a subgoal and a nal goal that weretemporally constrained may have posed a challenging task for children at this age. However,3-year-olds were successful in completing two noncontingent future goals in a planning task thatincluded few distracting elements and presented explicit goals (Experiment 3).

Both 3- and 4-year-olds were equally successful from both the self and other perspectives. The dis-tinction between planning from ones own perspective and planning from anothers perspective mayonly make a difference in future planning tasks that involve greater episodic content. For example,Russell and colleagues (2010) measured episodic future thinking by asking children to envision differ-ent spatial perspectives and found that 4-year-olds made more accurate choices when envisioning aspatial point of view from another childs perspective different from their own. In Russell and col-leagues study, it is possible that childrens original experience with the spatial environment couldhave interfered with their ability to imagine themselves in the same environment from a differentpoint of view. However, our study did not require spatial perspective taking and all of the locationswere in view at all times, and so children did not have the additional spatial demands evident in pastresearch (Russell et al., 2010; Suddendorf et al., 2011). In addition, in both the self and other condi-tions, children actively moved dolls through the spatial environment, further reducing the experientialdifferences in these conditions.

We measured basic future-oriented behavior through childrens ability to use past knowledge toconstruct possible future scenarios and achieve future goals. Our results suggest that early future-ori-ented behavior begins with semantic future thinking. We found that when goals were given in anexplicit form (Experiments 2 and 3), 3-year-olds were more accurate than when they needed to infera goal action from a mental state (Experiment 1). Furthermore, event construction of possible futurescenarios led to successful outcomes when goal complexity was reduced in Experiments 2 and 3. Inthe tasks presented here, children could have selected the correct location either by simulating out-comes for each location or by extracting knowledge from the past (i.e., where one should go to buya present and where one should go for Marys birthday). In both cases, a future-directed constructionprocess occurs and, furthermore, the process likely reects semantic processes. For this reason, pro-viding children with the goals they must accomplish allowed us to determine young childrens abilityto construct and enact goal-directed choices (and, therefore, future-directed choices). The additionaltask demands embedded in past methodologies where correct performance depended on recall andspatial perspective as well as episodic processes may have masked younger childrens ability to thinkabout the future (Russell et al., 2010; Suddendorf et al., 2011). By eliminating these extra taskdemands, simplifying the future environment, and having children enact actions toward specicfuture goals, our experiments elicited early signs of future thinking ability in 3-year-olds.

However, it is also important to consider the possibility that other cognitive processes may haveplayed a role in childrens performance in this study. Our tasks used future-oriented language andrequired children to keep multiple items and outcomes in mind at once. Specically, being able toaccomplish each goal required understanding the verbal instructions, and selecting an outcome likelyrequired sophisticated cognitive exibility and working memory abilities. However, because we didnot include these measures as covariates in our experiments, we cannot speak to the extent to whichlanguage and executive function ability play a role in accomplishing the goals. The experiments out-lined in this article reect initial explorations of 3- and 4-year-olds ability to maintain and accomplishgoals. Further research must be conducted to assess these individual components and their contribu-tion to future-oriented thinking.

Our study stands as a rst endeavor to track early future-oriented behavior that, over the preschoolyears, develops in complexity. We found that 3-year-olds were able to accomplish multiple goals,reecting an ability to use past knowledge to plan for future goals. This ability may underlie the devel-

J. Prabhakar, J.A. Hudson / Journal of Experimental Child Psychology xxx (2014) xxxxxx 13opment of prospective memory and other more complex uses of future thinking. We suggest that the


AlexandraEvideniere

AlexandraEvideniere

AlexandraEvideniere

AlexandraEvideniere

AlexandraEvideniere

AlexandraEvideniere

AlexandraEvideniere

AlexandraEvideniere

AlexandraEvideniere

AlexandraEvideniere

AlexandraEvideniere

AlexandraEvideniere

AlexandraEvideniereFUTURE DIRECTION

AlexandraEvideniere

AlexandraEvideniere

Hudson, J. A., & Mayhew, E. M. Y. (2011). Childrens temporal judgments for autobiographical past and future events. CognitiveDevelopment, 26, 331342.

Hudson, J. A., Mayhew, E. M. Y., & Prabhakar, J. (2011). The development of episodic foresight: Emerging concepts and methods.

Advances in Child Development and Behavior, 40, 95137.

Hudson, J. A. (1986). Memories are made of this: General event knowledge and development of autobiographic memory. In K.Nelson (Ed.), Event knowledge: Structure and function in development (pp. 97118). Hillsdale, NJ: Lawrence Erlbaum.

Hudson, J. A., Shapiro, L. R., & Sosa, B. B. (1995). Planning in the real world: Preschool childrens scripts and plans for familiarevents. Child Development, 66, 984998.

Kreitler, S., & Kreitler, H. (1987). Conceptions and processes of planning. In S. L. Friedman, E. K. Scholnick, & R. R. Cocking (Eds.),Blueprints for thinking: The role of planning in cognitive development (pp. 110178). Cambridge, MA: Cambridge UniversityPress.

Lemmon, K., & Moore, C. (2007). The development of prudence in the face of varying future rewards. Developmental Science, 10,502511.

McColgan, K. L., & McCormack, T. (2008). Searching and planning: Young childrens reasoning about past and future eventsequences. Child Development, 79, 14771497.

McCormack, K. L., & Hanley, M. (2011). Childrens reasoning about the temporal order of past and future events. CognitiveDevelopment, 266, 299314.

Metcalf, J. L., & Atance, C. M. (2011). Do preschoolers save to benet their future selves? Cognitive Development, 26, 371382.Moore, C., & Lemmon, K. (2001). The self in time: Developmental perspectives. Mahwah, NJ: Lawrence Erlbaum.Nelson, K., & Gruendel, J. (1986). Event knowledge: Structure and function in development. Hillsdale, NJ: Lawrence Erlbaum.Nelson, K., & Gruendel, J. (1981). Generalized event representations: Basic building blocks of cognitive development. In M. Lambbasic components of future thinking outlined in this article represent the basic requirements for pro-spective memory. To accomplish intended actions requires one to rst maintain a future goal and thento construct future event possibilities in which the intended action will take place. For instance, if oneneeds to remember to make a phone call during his or her lunch hour at work, one may constructstrategies for the future to help him or her remember to make the phone call. Basic goal maintenanceand construction abilities are required to ensure a successful plan and subsequent action to accom-plish the goal.

The ndings presented in this article indicate that the basic components of early future-orientedbehavior may underlie more complex future thinking behaviors that are evident in 3-year-olds.Throughout the preschool years, children further develop the ability to hold multiple goals and fea-tures in memory, which allows them to engage in more complex future thinking. These ndings addi-tionally suggest that the development of more complex future thinking occurs in tandemwith shifts inworking memory, event construction, and goal maintenance abilities.

References

Atance, C. M., & Jackson, L. (2009). The development and coherence of future-oriented behaviors during the preschool years.Journal of Experimental Child Psychology, 102, 379391.

Atance, C. M., & Meltzoff, A. N. (2005). My future self: Young childrens ability to anticipate and explain future states. CognitiveDevelopment, 20, 341361.

Atance, C. M., & ONeill, D. K. (2005a). The emergence of episodic future thinking in humans. Learning and Motivation, 36,126144.

Atance, C. M., & ONeill, D. K. (2005b). Preschoolers talk about future situations. First Language, 25, 518.Blewitt, P. (1982). Word meaning acquisition in young children: A review of theory and research. Advances in Child Development

and Behavior, 17, 139195.Busby, J., & Suddendorf, T. (2005). Recalling yesterday and predicting tomorrow. Cognitive Development, 20, 362372.DArgembeau, A., & Mathy, A. (2011). Tracking the construction of episodic future thoughts. Journal of Experimental Psychology:

General, 140, 258271.Einstein, G. O., & McDaniel, M. A. (1990). Normal aging and prospective memory. Journal of Experimental Psychology: Learning,

Memory, and Cognition, 16, 717726.Graf, P., & Uttl, B. (2001). Prospective memory: A new focus for research. Consciousness and Cognition, 10, 437450.Grant, J. B., & Suddendorf, T. (2010). Young childrens ability to distinguish past and future changes in physical and mental

states. British Journal of Developmental Psychology, 28, 8795.Grant, J. B., & Suddendorf, T. (2011). Production of temporal terms by 3-, 4-, and 5-year-old children. Early Childhood Research

Quarterly, 26, 8795.Harner, L. (1975). Yesterday and tomorrow: Development of early understanding of the terms. Developmental Psychology, 11,

864865.Harner, L. (1982). Immediacy and certainty: Factors in understanding future reference. Journal of Child Language, 9, 115124.Hayne, H., Gross, J., McNamee, S., Fitzgibbon, O., & Tustin, K. (2011). Episodic memory and episodic foresight in 3- and 5-year-

old children. Cognitive Development, 26, 343355.Hudson, J. A., & Fivush, R. (1991). As time goes by: Sixth graders remember a kindergarten experience. Applied Cognitive

Psychology, 5, 347360.

14 J. Prabhakar, J.A. Hudson / Journal of Experimental Child Psychology xxx (2014) xxxxxx& A. Brown (Eds.), Advances in developmental psychology (pp. 131158). Hillsdale, NJ: Lawrence Erlbaum.


AlexandraEvideniere

AlexandraEvideniere

AlexandraEvideniere

Nigro, G., Brandimonte, M. A., Cicogna, P., & Cosenza, M. (in press). Episodic future thinking as a predictor of childrensprospective memory. Journal of Experimental Child Psychology. doi: 10.1016/j.jecp.2013.10.013.

Raby, C. R., & Clayton, N. S. (2009). Prospective cognition in animals. Behavioral Processes, 80, 314324.Russell, J., Alexis, D., & Clayton, N. (2010). Episodic future thinking in 3- to 5-year-old children: The ability to think of what will

be needed from a different point of view. Cognition, 114, 5671.Schmider, E., Ziegler, M., Danay, E., Beyer, L., & Bhner, M. (2010). Is it really robust? Reinvestigating the robustness of ANOVA

against violations of the normal distribution assumption. Methodology: European Journal of Research Methods for theBehavioral and Social Sciences, 6, 147151.

Slackman, E. A., Hudson, J. A., & Fivush, R. (1986). Actions, actors, links, and goals: The structure of childrens eventrepresentations. In K. Nelson (Ed.), Event knowledge: Structure and function in development (pp. 4769). Hillsdale, NJ:Lawrence Erlbaum.

Suddendorf, T., & Corballis, M. C. (2007). The evolution of foresight: What is mental time travel and is it unique to humans?Behavioral and Brain Sciences, 30, 299313.

Suddendorf, T., Nielson, M., & von Gehlen, R. (2011). Childrens capacity to remember a novel problem and to secure its futuresolution. Developmental Science, 14, 2633.

Szpunar, K. K. (2010). Episodic future thought: An emerging concept. Perspective on Psychological Science, 5, 142162.Tulving, E. (2002). Episodic memory: From mind to brain. Annual Review of Psychology, 53, 125.

J. Prabhakar, J.A. Hudson / Journal of Experimental Child Psychology xxx (2014) xxxxxx 15Please cite this article in press as: Prabhakar, J., & Hudson, J. A. The development of future thinking: Young chil-drens ability to construct event sequences to achieve future goals. Journal of Experimental Child Psychology (2014),http://dx.doi.org/10.1016/j.jecp.2014.02.004

AlexandraEvideniere

AlexandraEvideniere

The development of future thinking: Young childrens ability to construct event sequences to achieve future goalsIntroductionExperiment 1MethodParticipantsMaterialsDesign and procedure

ResultsOverall accuracyResponses to subgoal questionResponses to final goal question

Discussion

Experiment 2MethodParticipants

Materials and procedureResults and discussion

Experiment 3MethodParticipantsMaterials and procedure


General discussionReferences

Documents

!!Development of Future Thinking in Young Children-semiopt-PrabhakarHudson, 2014