Preschoolers’ fantasy–reality distinctions of emotional events

Journal of Experimental Child Psychology 112 (2012) 467–483

Contents lists available at SciVerse ScienceDirect

Journal of Experimental ChildPsychology

journal homepage: www.elsevier .com/locate/ jecp

Preschoolers’ fantasy–reality distinctions of emotional events

Nathalie Carrick ⇑, Madisenne RamirezDepartment of Child and Adolescent Studies, California State University Fullerton, Fullerton, CA 92834, USA

a r t i c l e i n f o

Article history:Received 12 December 2011Revised 17 April 2012Available online 18 May 2012

Keywords:Children’s fantasy–reality distinctionsPretenseReality statusDiscrete emotionsEmotion regulationCognitive development

0022-0965/$ - see front matter � 2012 Elsevier Inhttp://dx.doi.org/10.1016/j.jecp.2012.04.010

⇑ Corresponding author. Fax: +1 657 278 4456.E-mail address: [email protected] (N. Carr

a b s t r a c t

Research suggests that emotions influence children’s ability to dis-cern fantasy from reality; however, reasons for this associationremain unknown. The current research sought to better understandthe mechanisms underlying children’s distinctions by examiningthe roles discrete emotions and context have in 3- to 5-year-olds’evaluations of fantasy and reality. In Study 1, children’s fantasy–reality distinctions of images depicting happy, frightening, or sadevents were obtained under two conditions: Children respondedfreely or were motivated to respond accurately. When respondingfreely, older children reported that happy and sad fantastic and realevents could occur but frightening events could not. Whenmotivated to respond accurately, children’s accuracy improved forfantasy but not for reality. Study 2 examined real events exclusivelyand found that children may lack the knowledge that frightening,but not happy or sad, real events can occur. The findings providenew insight into how emotions affect children’s fantasy–realitydistinctions.

� 2012 Elsevier Inc. All rights reserved.

Introduction

Children are routinely required to discern fantastic from real information. They must understandthat what they read in storybooks or see on television may or may not reflect reality, that events dur-ing pretend play are only make-believe, and that everyday conversations can include real or falseinformation. Typically, maneuvering between fantastic and real information does not prove to beoverly troublesome to young children. Research has examined several aspects of children’s evaluationand use of fantastic and real information, including understanding fictional or fantastic entities (e.g.,Skolnick & Bloom, 2006, 2009; Tullos & Woolley, 2009; Woolley & Cox, 2007), belief in imaginary

c. All rights reserved.

ick).

http://dx.doi.org/10.1016/j.jecp.2012.04.010

mailto:[email protected]

http://dx.doi.org/10.1016/j.jecp.2012.04.010

http://www.sciencedirect.com/science/journal/00220965

http://www.elsevier.com/locate/jecp

468 N. Carrick, M. Ramirez / Journal of Experimental Child Psychology 112 (2012) 467–483

companions (Taylor, Shawber, & Mannering, 2009), understanding of magic (Browne & Woolley, 2004;Subbotsky & Slater, 2011), and discerning between improbable and impossible events (Cook & Sobel,2011; Shtulman & Carey, 2007). Although age-related improvements exist across most studies,children typically treat information that represents reality differently from information that violatesreality, suggesting that they appreciate the boundaries of imaginary or fantastic information.

This conclusion, however, appears to be true primarily for information that is neutral rather thanemotional in tone. Studies have found that children’s competency at discerning fantasy and reality de-creases for emotionally evocative information. For example, when 4- to 7-year-olds are asked to imag-ine positive or negative entities (e.g., puppy, monster) inside empty boxes, they approach boxescontaining positive entities and avoid boxes containing negative entities, suggesting that the chil-dren’s emotional reaction to the entities increases their belief that the entities are real (Bourchier &Davis, 2000a, 2000b; Harris, Brown, Marriott, Whittall, & Harmer, 1991). Other research, however,has found that when 3- to 5-year-olds are asked to evaluate images of fantastic or real events thatare frightening, they will accurately report that frightening fantastic events are not real (Sayfan &Lagattuta, 2009) but inaccurately report that frightening real events are not real (Samuels & Taylor,1994), suggesting that fear can decrease children’s belief in fantasy and elicit errors in reality. Takentogether, these studies reveal that emotions influence children’s evaluations of fantastic and realinformation but do not clearly document whether emotions increase or decrease the accuracy ofchildren’s judgments or provide reasons for the effect of emotion.

One study attempted to reconcile prior findings by systematically testing children’s fantasy–realitydistinctions of events depicting several discrete emotions: happiness, fear, and anger (Carrick & Quas,2006). Discrete emotions were selected because they vary in both valence (i.e., positive vs. negative)and motivational tendencies (i.e., approach vs. avoidance), which can provide insight into children’sreactions to and judgments of the information. That is, children viewing emotional events associatedwith approach tendencies may find them to be desirable and want to engage with the events, hencereporting that the events are real, versus events associated with avoidant tendencies, which childrenmay want to disengage from and report are not real. For example, happiness is positively valenced andassociated with approach tendencies (Carver & White, 1994; Gray, 1981, 1990); therefore, individualsshould report that positive events (fantastic or real) are real in order to maintain positive feelings orengage with the events. Fear is negatively valenced and associated with avoidant tendencies (e.g.,Davidson, 1998; Lazarus, 1991; Wacker, Heldmann, & Stemmler, 2003); thus, individuals should re-port that frightening events are not real in order to avoid or distance themselves from the events. An-ger is negatively valenced like fear, but it is associated with individuals engaging with the anger-eliciting stimulus to eliminate a blocked goal (Berkowitz, 1999; Harmon-Jones & Sigelman, 2001; Laz-arus, 1991; Levine, 1995; Wacker et al., 2003). Therefore, judgments can vary depending on whetherindividuals are responding to the negative valence (and report that events cannot occur) or approachtendency (and report that events can occur).

Carrick and Quas (2006) found that when 3- to 5-year-olds viewed fantastic and real happy, fright-ening, and angry events and were asked, ‘‘Can this happen in real life?’’ children reported that happyevents (e.g., mice dancing, parent hugging child) could occur more often than frightening events(e.g., witch, house on fire) and angry events (e.g., raccoons arguing, parents yelling at child) regardlessof whether events depicted fantasy or reality. Thus, children did not discriminate the fantastic or realstatus of the images or the motivational tendency (approach vs. avoidance); rather, they differentiallyjudged the images based on the positive or negative valence of the events. The authors suggested thatchildren’s responses reflect their desires for positive but not negative events to occur or children’s at-tempts to regulate their positive or negative reactions to viewing the events. Although these findingbegin to capture the association between emotions and children’s fantasy–reality distinctions, twoadditional issues must be resolved before a thorough understanding of this association can be obtained.

First, examining the contribution of discrete emotions versus general emotional valence to chil-dren’s judgments may better explain why children report that certain events can or cannot occur.In Carrick and Quas’s (2006) study, children reported that the angry events could not occur, suggestingthat children responded to the emotional valence more so than the motivational tendency associatedwith anger. However, before concluding that valence is the primary influence, other discrete negativeemotions need to be examined. Sadness is a useful emotion to test because it can differentially affect a

N. Carrick, M. Ramirez / Journal of Experimental Child Psychology 112 (2012) 467–483 469

person depending on the context surrounding the sad event and the intensity of sadness a personfeels. Traditional appraisal theory argues that sadness elicits avoidant tendencies because a personrealizes that a goal is lost and, in turn, disengages with the event (e.g., Harmon-Jones & Allen,1998; Lazarus, 1991; Levine, 1995; Roseman, 1991; Stein & Levine, 1999). However, when witnessinga sad event that is not overly distressing, children have been found to respond with empatheticbehaviors (e.g., offer a toy to a crying friend) (Eisenberg & Fabes, 1992; Eisenberg et al., 1994; Fabes,Eisenberg, McCormick, & Wilson, 1988; Guthrie et al., 1997). Related to the current research, whenchildren view sad events that do not overly distress them, they may report that these events can occuras an attempt to engage with the characters. Thus, in the current studies, images depicting sad events,along with frightening and happy events, are included to test whether discrete emotions provide a un-ique contribution to children’s fantasy–reality distinctions or, rather, children base their judgmentssolely on emotional valence.

The second issue concerns how the context surrounding children’s evaluations of fantasy and real-ity influences their fantasy–reality distinctions. Research suggests that children’s (and even adults’)evaluation of fantastic and magical information is heavily influenced by context. Children’s endorse-ment of fantasy may increase when it is associated with positive outcomes (e.g., believing in SantaClaus is associated with receiving a gift) or may decrease when it is tied to negative outcomes (e.g.,adult disapproval) (Taylor & Carlson, 2000; Woolley & Phelps, 1994). Adults and children have evenshown endorsement of magical thinking to avoid negative consequences (Subbotsky, 1994, 2001).Woolley, Boerger, and Markman (2004) argued that when evaluating the reality status of information,children weigh the costs versus benefits of endorsing fantasy and base their judgments on howendorsing fantasy will affect them. For example, the authors found that 4-year-olds’ belief in a novelfantastic entity (Candy Witch) that exchanges children’s Halloween candy for presents increased forthose children who preferred presents to candy versus those children who preferred candy. For thechildren who preferred presents, endorsement of the Candy Witch benefited them, unlike thosechildren who preferred candy, for which endorsement of the Candy Witch was too great a cost.

Thus, when reconsidering Carrick and Quas’s (2006) findings in this perspective, children’sresponses that happy events can occur but frightening and angry events cannot occur may have beendue to the context surrounding their judgments. That is, children were not told that they needed toprovide accurate fantasy–reality distinctions; thus, they responded how they wished—positive eventscan occur and negative events cannot. There were no costs associated with children’s errors of endors-ing positive fantasy and denying negative reality. In fact, denying that negative reality can occur mayhave benefited children by allowing them to regulate any negative feelings associated with the fright-ening or angry events. Thus, children’s fantasy–reality distinctions may have been a result of the test-ing context, leaving their actual beliefs still unknown. To gain a better understanding of their actualbeliefs, the testing context in Study 1 was manipulated to examine whether children’s fantasy–realitydistinctions improved when presented with a benefit to provide accurate responses (i.e., children wererewarded with a prize).

The current studies were designed to address the issues of discrete emotions and context to estab-lish a more thorough understanding of the relation between emotion and children’s fantasy–realitydistinctions. In Study 1, 3- to 5-year-olds’ fantasy–reality distinctions for happy, frightening, andsad events were examined under two contexts. The first context was similar to the methods employedin Carrick and Quas (2006), such that children rendered their fantasy–reality distinctions without anyrestriction or manipulation. In the second context, however, the same children viewed new stimuliand were rewarded for correct fantasy–reality distinctions. Because results revealed that rewardingchildren improved their accuracy for fantastic but not real events, Study 2 was designed to examinereal events exclusively and the mechanisms underlying children’s evaluations of emotional reality.

Study 1

Study 1 had two main goals. The first goal was to expand the discrete emotions included in priorstudies (e.g., Carrick & Quas, 2006) to include sadness along with happiness and fear. As stated,sadness was included because although it is negatively valenced like fear, it may activate approach


tendencies in that when viewing sad events, children may have an empathetic response to the char-acters. The second goal was to test for differences in children’s fantasy–reality distinctions due to con-textual differences, specifically when children respond freely or are rewarded for correct responses.Children viewed two sets of images depicting fantastic and real events that showed happiness, fear,or sadness. During the first set of images, no restriction was placed on children’s fantasy–reality dis-tinctions, but during the second set of images, children were told that they would receive a prize ifthey provided accurate fantasy–reality distinctions. It was predicted that during the first set of images,children would report that the happy and sad events could occur more often than the frighteningevents regardless of fantasy or reality. During the second set of images, children were expected to re-port more accurately that all fantastic events cannot occur and real events can.

Methods

ParticipantsThe participants were 92 3- to 5 year-olds (M = 54 months, 49% female and 51% male): 30 3-year-

olds (M = 44 months), 34 4-year-olds (M = 54 months), and 28 5-year-olds (M = 64 months). The sam-ple was ethnically diverse (59% Caucasian, 5% African American, 21% Latino, and 15% Asian American)and recruited from child-care centers across diverse socioeconomic communities. An additional 9children asked to stop the testing session midway and were not included in the analyses.

MaterialsFantasy–reality images were selected from children’s storybooks and eligible for inclusion if they

depicted fantastic or real events, included only one emotion, were not cartoons, and did not depictpopular images known to children (e.g., Santa Claus). Undergraduate students (N = 95) viewed 50images and rated them on the likelihood of each event occurring (1 = definitely not to 5 = definitelyyes), how the image made them feel (1 = really really bad to 5 = really really good), and which emotionwas depicted. Mean likelihood scores for the fantastic images ranged from 1.39 to 1.68 and for the realimages ranged from 4.04 to 4.86. Mean emotional ratings for images were 4.11 for happy, 2.36 forfrightening, and 2.21 for sad. Percentages of students who correctly identified the emotion were93% for happy, 82% for frightening, and 70% for sad.

A total of 36 images that received the highest ratings were selected: 18 fantasy and 18 reality, withan equal number of images depicting happiness, fear, and sadness. See Appendix A for a description ofthe images. To confirm that children could accurately identify the discrete emotion in each image, 3-to 5-year-olds (N = 14) viewed each image and were asked, ‘‘What do you see in the image?’’ and‘‘How does he/she [pointing to the character in the image] feel?’’ and then pointed to a face on a scaledepicting four emotional states: neutral, happy, frightening, and sad (Quas, Wallin, Horwitz, Davis, &Lyon, 2009). The percentages of children who correctly labeled the character’s feelings for each emo-tion were 93% for happy, 72% for frightening (72%), and 94% for sad.

ProcedureThe procedure was approved by California State University Fullerton’s internal review board, and

parental consent was obtained prior to testing. Children were tested individually in a quiet room attheir child-care center.

Phase 1. Children viewed four practice images (two fantasy and two reality) and were asked three testquestions: (a) ‘‘What do you see in this picture?’’ (this question served to ensure that children at-tended to the image); (b) ‘‘How does this picture make you feel?’’ (accompanied by a face scale depict-ing five faces, ranging from a large smile to a large frown, associated with really really good, good, okay,bad, and really really bad [Carrick & Quas, 2006; Carrick, Quas, & Lyon, 2010]); (c) ‘‘Can this happen inreal life?’’ (errors on this question were corrected during the practice images).

After the practice images, children were administered 18 test images (half fantasy and half reality,with an equal number of each emotion). Children viewed each image separately and answered thethree test questions. Children’s responses were not corrected. Children who responded with ‘‘I don’tknow’’ were encouraged to choose either ‘‘yes’’ or ‘‘no’’.


Phase 2. After a short break, children began Phase 2. Before viewing the new set of 18 test images,children were told, ‘‘Now I’m going to show you some different pictures and ask you the same ques-tions. But this time, it will be different. If you get all the answers right, you’ll get a prize.’’ The research-er then presented the children with a gift bag that contained a prize and said, ‘‘So if you get all theanswers right, you’ll get the prize in this bag. So you have to think really hard and give the right an-swer to all the pictures.’’ The bag remained close to the children throughout the phase as a visualreminder.

Children were given four practice images (two fantasy and two reality) and told that if they werecorrect on the four practice images, they would receive a sticker. This served as a demonstration thatchildren would receive a prize for correct answers. Children were asked the three test questions. Forcorrect responses to the question, ‘‘Can this happen in real life?’’ the researcher told the children verycheerfully that they were correct. For incorrect responses, the researcher told the children that theywere wrong, gave them the correct answer, and asked the question again until the children answeredcorrectly. Children were given the sticker after the four practice items, and it was placed next to thegift bag. Children were then shown the test images and asked the three test questions. Incorrectresponses were not corrected. Every third image, the researcher drew the children’s attention to theprize bag and reminded them that they needed to answer correctly to get the prize. At the end, allchildren were given the prize.

The order of the images within each phase was randomized with the exception that the last imagedepicted a happy event to alleviate any negative feelings the frightening or sad images may haveevoked. In addition, images were counterbalanced across phases to reduce possible effects onchildren’s performance; six versions were created, so that each image appeared an equal number oftimes in Phase 1 and Phase 2.

ScoringChildren’s responses to the question, ‘‘Can this happen in real life?’’ were scored as 0 for incorrect

responses and as 1 for correct responses. Mean scores for the fantastic and real events were calculatedfor each emotion and according to phase, creating a total of six scores per phase.

Results

Preliminary analyses revealed that ethnicity, gender, and order of images were unrelated to chil-dren’s performance and are not considered further. To ensure that children’s emotional responsesto the images differed between the positive and negative events, their ratings to the question,‘‘How does this make you feel?’’ were analyzed (5 = more negative rating, 1 = more positive rating).Scores were entered into a 3 (Age: 3, 4, or 5 years) � 2 (Type of Image: fantasy or reality) � 3 (Emo-tion: happy, frightening, or sad) repeated-measures analysis of variance (ANOVA).1 A main effect foremotion, F(2,170) = 78.02, p < .001, g2

p ¼ :47, revealed that happy images (M = 1.88) were rated as morepositive than frightening (M = 3.43, p < .001) and sad (M = 3.21, p < .001) images, and frightening and sadimages were rated as comparably negative. Thus, children’s emotional ratings differed between positiveand negative events.

Phase 1Phase 1 focused on differences in children’s fantasy–reality distinctions for happy, frightening, and

sad events. Children were expected to report that happy and sad events could occur more often thanfrightening events regardless of fantasy or reality. Children’s distinction scores were entered into arepeated-measures ANOVA with age (3, 4, or 5 years) as the between-participants measure and imagetype (fantasy or reality) and emotion (happiness, fear, or sadness) as the within-participants mea-sures. A main effect of age, F(2,89) = 13.46, p < .001, g2

p ¼ :23, revealed that 5-year-olds were moreaccurate on fantasy–reality distinctions than 3-year-olds (p < .001) and 4-year-olds (p = .004), whodid not differ from each other. A main effect of emotion also emerged, F(2,178) = 9.39, p < .001,

1 For repeated measures analyses, Huynh–Feldt statistics are reported because sphericity could not be assumed.


g2p ¼ :09. Children were more accurate on happy images than on frightening (p < .001) and sad (p = .05)

images, with no difference between the latter two. However, these effects were subsumed by asignificant Image Type � Emotion interaction, F(2,178) = 45.69, p < .001, g2

p ¼ :33, and an ImageType � Emotion � Age interaction, F(4,178) = 2.87, p = .02, g2

p ¼ :06. To explore the three-way interac-tion, repeated-measures ANOVAs were conducted on children’s fantasy–reality distinction scores sep-arately by age. For each age group, the interaction between image type and emotion was significant;thus, repeated-measures ANOVAs were conducted separately for fantasy images and reality images byage. See Table 1 for means.

For 3-year-olds, the main effect of emotion for fantastic images approached significance,F(2,58) = 3.04, p = .056, g2

p ¼ :09. Children reported that fantastic happy events could occur more thanfantastic frightening events (p = .032), and no differences emerged between frightening and sadevents. The main effect of emotion for real images was significant, F(2,58) = 12.65, p < .001,g2

p ¼ :30, revealing that 3-year-olds reported that real happy events could occur more often than realfrightening (p = .001) and sad (p = .006) events, with the latter two not differing from each other.

For 4- and 5-year-olds, significant differences among emotions emerged for both fantastic and realimages, Fs(2,54) P 3.39, ps < .04, g2

p ¼ :30values � :11. That is, across fantasy and reality, these chil-dren reported that happy and sad events could occur more often than frightening events (ps < .02).Their judgments of happy and sad images did not differ. One exception emerged for 5-year-olds,who did not significantly report that fantastic happy events could occur more often than fantasticfrightening events.

These analyses provided partial support for the study’s hypotheses. The 3-year-olds showed a re-sponse bias to report that all happy events could occur more often than frightening events and realhappy events could occur more often than real sad events, suggesting that their judgments were basedmore on valence than on discrete emotion. The 4- and 5-year-olds, in contrast, were more likely toreport that happy and sad events could occur more often than frightening events, suggesting that old-er children differentially based their judgments on discrete emotions.

Phase 2Phase 2 examined changes in children’s accuracy when presented with a prize for correct

fantasy–reality distinctions. Children were expected to provide more accurate responses during Phase2 compared with Phase 1. Children’s fantasy–reality distinction scores were entered into a repeated-measures ANOVA separately for fantastic and real events. Age served as the between-participants mea-sure, and phase (1 or 2) and emotion served as the within-participants measures. See Table 1 for means.

Table 1Study 1 means and standard deviations of fantasy–reality distinction scores across emotion and phase.

3-year-olds 4-year-olds 5-year-olds

Phase 1 Phase 2 Phase 1 Phase 2 Phase 1 Phase 2

Fantastic eventsHappy .52 (.40)a .63 (.40) .62 (.43)a .65 (.40)⁄ .82 (.29)ab⁄ .92 (.22)⁄

Frightening .68 (.41)b⁄ .79 (.36)⁄ .82 (.26)b⁄ .81 (.34)⁄ .90 (.24)b⁄ .87 (.28)⁄

Sad .59 (.39)ab .69 (.38)⁄ .70 (.36)a⁄ .80 (.35)⁄ .73 (.35)a⁄ .86 (.29)⁄

Mean .60 (.40)a .69 (.38)b .69 (.36)a .75 (.36)b .81 (.29)a .89 (.24)b

Real eventsHappy .84 (.27)a⁄ .86 (.27)⁄ .88 (.26)a⁄ .99 (.06)⁄ .94 (.22)a⁄ .98 (.13)⁄

Frightening .51 (.44)b .39 (.37) .45 (.39)b .43 (.41) .63 (.34)b⁄ .77 (.29)⁄

Sad .61 (.42)b .59 (.44) .74 (.37)a⁄ .73 (.38)⁄ .93 (.16)a⁄ .96 (.10)⁄

Mean .65 (.39)a .65 (.36)a .73 (.32)a .76 (.26)a .86 (.22)a .91 (.18)a

Note: Standard deviations are in parentheses. Higher fantasy–reality distinction scores indicate a greater proportion of accurateresponses that fantastic events cannot occur and real events can occur. Superscripts denote significant differences in scoreswithin each age group for Phase 1. Subscripts denote significant differences in scores within each age group across emotion andphase. Asterisk (�) denotes means that are significantly different from chance at p 6 .05.


For fantastic events, only the main effect of phase was significant, F(1,89) = 6.83, p = .011, g2p ¼ :07,

revealing that across age and emotions, children were more accurate in reporting that fantasy couldnot occur during Phase 2 relative to Phase 1. In contrast, for real events, no significant main effectsor interactions emerged, revealing that the manipulation did not influence children’s judgments ofreality.

Of note, children’s ratings of how the images made them feel were analyzed to test whetherincreases in children’s accuracy during Phase 2 may have been due to changes in their intensity ofemotional ratings to the images. Scores were entered into a repeated-measures ANOVA with age asthe between-participants measure and phase and emotion as the within-participants measures. Onlya significant main effect of emotion emerged, F(2,160) = 106.19, p < .001, g2

p ¼ :57. Across phases,children rated happy events (M = 1.85) as more positive than frightening (M = 3.47, p < .001) andsad (M = 3.39, p < .001) events, and they rated frightening and sad events as comparably negative.Importantly, no differences in children’s emotional ratings emerged across phase. Thus, not only werechildren correctly discriminating among positive and negative emotions during Phase 2, but also anyincreases in accuracy could not be attributed to changes in children’s emotional reactions to theimages. Instead, it appears that the testing condition associated with Phase 2 had a positive influenceon children’s accuracy for fantasy but had no effect on children’s accuracy for reality.

Discussion

Study 1 revealed age-related changes in how emotional valence and discrete emotions influencechildren’s fantasy–reality distinctions during Phase 1. The 3-year-olds judged that fantastic and realhappy events could occur more often than frightening events and that real happy events could occurmore often than real sad events, suggesting that their judgments were based on valence versus dis-crete emotions. In contrast, discrete emotions appeared to affect the 4- and 5-year-olds’ judgments.These age groups were more likely to report that happy and sad events could occur more often thanfrightening events. It is somewhat surprising that the 3-year-olds did not show the same response pat-tern for sadness and fear as the older children given that they can differentiate between sadness andfear (Levine, 1995; Stein & Levine, 1999; Widen & Russell, 2010) and signs of offering condolence havebeen found in children as young as 2 years (e.g., Zahn-Waxler, Radke-Yarrow, Wagner, & Chapman,1992). Perhaps this fantasy–reality distinction task placed a high cognitive demand on young childrenand they could not integrate their feelings of concern for others with their fantasy–reality judgments.

Prior research found differences in fantasy–reality distinctions based simply on valence (Bourchier& Davis, 2000a, 2000b; Carrick & Quas, 2006; Harris et al., 1991; Samuels & Taylor, 1994), but morecomplex processes appear to influence older children’s judgments more than simply avoiding negativeinformation and engaging with positive information. Older children’s willingness to report that sadevents can occur more often than frightening events may reflect their perception of sadness as lesspersonally threatening than fear or their greater resources to cope with sadness. The sad eventsmay also have elicited feelings of concern or empathy (e.g., Fabes et al., 1988; Hoffman, 1985). In con-trast, fear presents a potential threat that warrants avoidance. This is especially likely if children inter-preted the test question, ‘‘Can this happen in real life?’’ as ‘‘Can this happen to me?’’ Their responsesmay represent a coping strategy to reduce the negative feelings associated with the event (e.g., ‘‘Scarythings aren’t real’’) (Sayfan & Lagattuta, 2009). Alternatively, because parents often try to shelterchildren from frightening information, children may have less exposure to frightening events and lackthe knowledge that they can occur. Parents may also try to lessen the effect that frightening eventshave on children by describing the events as unreal (Fivush, Berlin, Sales, Mennuti-Washburn, & Cas-sidy, 2003). In fact, in a study testing fantasy–reality distinctions among typically developing childrenand maltreated children who had been routinely exposed to negative events, Carrick et al. (2010)found that typically developing children were less likely than maltreated children to report that fright-ening events could occur. These findings suggest that perhaps children’s experiences and knowledgeof negative events occurring may influence their evaluations of fear. However, Study 1 was not able totease apart whether children’s responses reflect a lack of knowledge or a strategy to cope with nega-tive emotions. Study 2, therefore, was designed to test these two explanations.


Children’s performance during Phase 2, in which they received a prize for correct responses,revealed that the testing context influenced children’s judgments of fantasy but not of reality. Whenchildren were motivated to respond correctly, they were significantly better able to report that fantas-tic events could not occur, suggesting that children have a basic understanding that fantastic eventsare not real but may endorse fantasy when the context permits it. In terms of weighing the costsand benefits of their responses, there was no true cost to reporting that events such as happy dancingmice are real during Phase 1. But when a consequence was introduced during Phase 2 and childrenneeded to choose between responding how they would like and responding correctly to receive aprize, their perceived benefit shifted to the prize. Such a pattern is consistent with prior work(Woolley & Phelps, 1994; Woolley et al., 2004); however, the current study is unique in that it system-atically examined a range of emotional events and found that children’s reports of happy, frightening,and sad fantasy alike improved.

The manipulation during Phase 2 did not, however, increase children’s accuracy for real events. Forhappy images, this is unsurprising given children’s near ceiling accuracy during Phase 1. In addition,for frightening images, a lack of increase is consistent with the proposed explanations for why childrenreported that frightening events cannot occur during Phase 1. Either the prize was not enough moti-vation for children to override any threatening feelings elicited by the images, or children simply donot think that these events can occur and no prize can change this belief. Yet, why children’s accuracyfor sad events did not increase remains somewhat unclear. Accuracy for sad images during Phase 1was quite high; however, there were still children who inaccurately reported that sad events cannotoccur. Perhaps these children did not increase accuracy during Phase 2 because they did not know thatsad events can occur in a manner similar to frightening events. This conclusion cannot yet be madebecause Study 1 did not examine underlying reasons for children’s judgments. Study 2, therefore,was designed to provide explanations for children’s evaluations of real events.

Another limitation of Study 1 concerns the images employed. Images varied by discrete emotionbut also varied in content (i.e., the scene of the event). Content ranged from children playing, to micedancing, to monsters scaring people. Therefore, children’s judgments were based on more than onlyemotion. Holding the content of the images constant while depicting different emotions would allowmore robust conclusions regarding the influence of emotion. As such, Study 2 used images thatdepicted similar content and varied only in emotion.

In sum, Study 1 found that individual discrete emotions uniquely affect older children’s evaluationsof fantasy and reality and provided support for the notion that children’s understanding of fantasy isinfluenced by context. That is, children may appear to understand that fantasy is not real but reportthat it is so under certain situations. However, different processes are likely at play for evaluationsof reality insofar as children’s accuracy of reality did not increase during Phase 2. This possibility isin need of further testing and was the main focus of Study 2.

Study 2

The purpose of Study 2 was to further examine children’s judgments of real events to provide anexplanation for Study 1’s finding that children reported that happy and sad real events can occur moreoften than frightening events. Do children’s reports correspond to their knowledge about the possibil-ity of events occurring, or are they attempts to cope with emotions? In Study 2, children viewedimages depicting happy, frightening, or sad real events and sorted them into one of three categories:events that could happen (a) to themselves (Self), (b) only to another child (Other), or (c) Never. Thecategories were designed to contrast children’s knowledge versus coping strategies. That is, selectingSelf should indicate that children know that an event can occur and accept that it can occur to them,selecting Other should represent knowing that an event can occur but preferring that it occur toanother person rather than themselves, and selecting Never should correspond to not knowing thatthe event can occur.

Of note, the real events employed in the study (e.g., a child and mother making cookies) can occurto both the Self and Other categories; therefore, both categories are correct. However, requiring chil-dren to choose one category tested whether they accepted that the event could occur to them rather


than preferring that it occur to another person. In addition, it is possible that children would selectNever for events that they do not want to occur versus do not know can occur. However, given theavailability of the Other category, which is more suited for this belief, selecting Never is more likelya sign of not knowing that the event can occur.

The following predictions were made. Based on the findings from Study 1, children were expectedto select Self for happy events because they are routinely exposed to these nonthreatening events andshould accept that they can occur to themselves. Older children were expected to select Other for sadevents given older children’s responses that sad events could occur in Study 1 but also the likelihoodthat they would prefer the events occur to someone else rather than to themselves. Older childrencould also select Other for frightening events if they know the events can occur but do not want themto occur to themselves. However, if they lack the knowledge that frightening events can occur, theyshould select Never. Therefore, whether children select Other or Never for frightening events will helpto explain why they reported that these events cannot occur in Study 1. Finally, predictions for whichevents the 3-year-olds would select for Other and Never remained exploratory given that theirresponses differed from those of older children in Study 1.

Methods

ParticipantsThe participants were 65 3- to 5-year-olds (M = 53 months, 53% female and 47% male): 22 3-year-

olds (M = 43 months), 21 4-year-olds (M = 53 months), and 22 5-year-olds (M = 62 months). The sam-ple was ethnically diverse (38% Caucasian, 5% African American, 24% Latino, 6% Asian American, and27% mixed ethnicity or ‘‘other’’) and recruited from child-care centers that serve economically diversefamilies. None of the children in Study 2 participated in Study 1. An additional 5 children wereincluded but asked to stop testing midway through the session and are not included in the analyses.

MaterialsTest images. A total of 18 images depicting real events were drawn by an artist. The images depictedsix different themes (e.g., child and house, child and parent baking, child and dog). Within each theme,3 images depicted happy, frightening, and sad versions of the theme. For example, in the child andhouse theme, the happy version depicted a smiling child standing outside the house, the sad versiondepicted a child crying outside the house, and the frightening version depicted a scared child outsidethe house on fire. A vignette describing the event accompanied each image. See Appendix B fordescriptions of images and vignettes.

The 18 images were shown to undergraduate students (N = 20) for their ratings on three dimen-sions: likelihood of event occurring (1 = definitely not to 5 = definitely yes), how the image made themfeel (1 = really really bad to 5 = really really good), and which emotion was depicted. Mean likelihoodscores ranged from 3.33 to 5. Mean emotional ratings for images were 4.0 for happy, 1.8 for frighten-ing, and 2.2 for sad. Percentages of students who correctly identified the emotion depicted in theimages were 100% for happy, 91% for frightening, and 91% for sad. Four themes with the highest rat-ings were selected for the final sample of images. Consistent with Study 1, the final set of images wasshown to 3- to 5-year-olds (N = 12). The percentages of children who correctly identified the charac-ter’s feelings for each emotion were 83% for happy, 72% for frightening, and 80% for sadness.

Box sorting task. Three boxes (9 � 11 inches, painted black) were created for the study. The Self boxhad a picture of the child (taken with a Polaroid camera) along with his or her name. The Otherbox had a picture of a boy or girl (depending on the child’s gender) with the label ‘‘Sam’’ or ‘‘Mary’’.The Never box had the label ‘‘Not Real’’ with an ‘‘X’’. The boxes had a thin slit in the front wherechildren slid in the images. After sliding in an image, it was no longer visible in order to prevent chil-dren from choosing a box based on prior choices.

ProcedureThe procedure was approved by the university’s internal review board, and parental consent was

obtained prior to testing. Children were tested individually in a quiet area of their child-care center.


The researcher placed the three boxes in front of the children and gave the following instructions: ‘‘I’mgoing to show you some pictures and ask you questions about the pictures. We are going to use theseboxes to answer the question. This box has your name written on it. And this box has a picture of agirl/boy named Sam/Mary with his/her name on it. Sam/Mary is a [age of child]-year-old just likeyou, but Sam/Mary lives very very far away from here in another city. You don’t know Sam/Maryand probably will never meet him/her. This box says Not Real and there’s an X. This means that thingsin this box can’t happen in real life; they are not real.’’ The researcher then presented the images andsaid, ‘‘These pictures show kids doing different things. Some are things that can happen to you, orthings that can only happen to another kid like Sam/Mary, or things that aren’t real and can’t happenin real life. I want you to look at the pictures and choose if what you see is something that can happento you, or only another kid, or never and then put it in the box you think is best.’’

Children were administered five practice images that depicted either real or impossible events andwere asked the two test questions: ‘‘How does this make you feel?’’ along with the same face scale asStudy 1 and ‘‘Can this happen to you, only another child, or never?’’ Children were corrected if theyplaced a real event in the Never box but not if they placed a real event in the Self or Other box. Thetest images were then administered. For each image, children were read the vignette and asked thetest questions. Children’s errors were not corrected. If they said, ‘‘I don’t know,’’ they were encouragedto select a box. Consistent with Study 1, the order of the images was randomized with the exceptionthat the last image was positive. The position of each box was counterbalanced across children.

ScoringThe box that children selected for each image served as their box selection score. A score of 1 was

given for selecting a box, and a score of 0 was given for not selecting a box. Mean scores were calcu-lated separately by emotion and box for a total of nine scores (e.g., Self box happy, Other boxfrightening).

Results

Preliminary analyses revealed no association between children’s gender, ethnicity, order of images,or boxes and their box selection scores; thus, none is considered further. Consistent with Study 1,children’s ratings in response to the question, ‘‘How does this make you feel?’’ were analyzed toensure that children’s emotional reactions were different for the positive and negative events(5 = more negative rating, 1 = more positive rating). Children’s scores were entered into a 3 (Age: 3, 4,

Table 2Study 2 means and standard deviations of children’s box selection scores across emotion and box.

3-year-olds 4-year-olds 5-year-olds

Self boxHappy .43 (.40)a .70 (.38)a⁄ .68 (.26)a⁄

Frightening .22 (.34)b .21 (.36)b .26 (.34)b

Sad .24 (.37)b .14 (.25)b⁄ .30 (.37)b

Other boxHappy .37 (.42)a .21 (.36)a .17 (.25)a⁄

Frightening .41 (.42)a .26 (.30)a .17 (.29)a⁄

Sad .33 (.38)a .48 (.37)b .32 (.35)b�

Never boxHappy .21 (.31)a .08 (.15)a⁄ .15 (.20)a⁄

Frightening .37 (.41)b .53 (.36)b⁄ .58 (.37)b⁄

Sad .43 (.40)b .37 (.35)a� .38 (.38)c

Note: Standard deviations are in parentheses. Higher box selection scores indicate greater pro-portion of selecting a box by emotion. Superscripts denote significant differences in scores withineach age group per box (�p = .07). Asterisk (�) denotes means that are significantly different fromchance at p 6 .05.


or 5 years) � 3 (Emotion: happy, frightening, or sad) repeated-measures ANOVA. A main effect foremotion, F(2,122) = 169.78, p < .001, g2

p ¼ :74, revealed that happy images (M = 1.58) were rated asmore positive than frightening (M = 4.20, p < .001) and sad (M = 4.19, p < .001) images, and frighteningand sad images were rated as comparably negative. Thus, children’s emotional ratings differed be-tween positive and negative events.

Children’s box selectionIt was hypothesized that children’s judgments of whether events could occur to the Self, Other, or

Never would depend on the emotion depicted. Children’s box selection scores were entered into arepeated-measures ANOVA with age (3, 4, or 5 years) as the between-participants measure and box(Self, Other, or Never) and emotion (happy, frightening, or sad) as the within-participants measures.A significant Box � Emotion interaction emerged, F(4,244) = 25.30, p < .001, g2

p ¼ :29, but was sub-sumed under a Box � Emotion � Age interaction, F(8,244) = 2.613, p = .009, g2

p ¼ :08. To explain thisinteraction, subsequent simple effects analyses were conducted, and results are discussed separatelyfor each category: Self, Other, and Never. See Table 2 for means.

Selecting the Self box was hypothesized to represent events children know and accept can occur tothemselves. Children were expected to select happy events more often than frightening or sad eventsfor the Self box. Results revealed a main effect of emotion for each age group (ps 6 .031); children se-lected the Self box for happy events more often than for frightening and sad events, with the caveatthat the difference between happy and sad events for 3-year-olds was a trend (p = .06). Mean scoresbetween frightening and sad events did not differ.

Selecting the Other box was hypothesized to represent knowing an event can occur but preferringthat it occur to another person. Older children were expected to show a preference to place sad eventsin the Other box. Results reveal age differences for which events children placed in the Other box. The3-year-olds placed happy, frightening, and sad events equally in the Other box, F(2,40) = 0.531, p > .05.The 4-year-olds showed a selection bias based on emotion, F(2,40) = 4.62, p = .016, g2

p ¼ :18 andplaced sad events into the Other box significantly more often than happy (p = .01) and frightening(p = .02) events, with no difference between happy and frightening events. For the 5-year-olds,although the pattern of selection was similar to that for the 4-year-olds, the main effect of emotiononly approached significance, F(2,42) = 2.692, p = .079. Therefore, older children provided some evi-dence that they may understand that sad events can occur but do not want them to occur tothemselves.

The Never box, in contrast, was hypothesized to represent events that children believe could notoccur. Older children were expected to select frightening events more often than happy and sad eventsfor the Never box. The findings showed that children’s selections for the Never box varied by age. Amain effect of emotion emerged for the 3-year-olds, F(2,40) = 5.389, p = .008, g2

p ¼ :21, who selectedthe Never box more often for frightening and sad events than for happy events (ps 6 .047), suggestingan influence of valence versus discrete emotion (consistent with Study 1). Main effects of emotion alsoemerged for the 4- and 5-year-olds, Fs > 13.63, ps < .001, g2

pvalues > :39, but their selection patternwas different from that of the 3-year-olds. The 4-year-olds selected the Never box for frighteningevents more often than for happy events (p < .001) and somewhat more often than for sad events(p = .076). The 5-year-olds picked Never for frightening events more often than for happy (p < .001)and sad (p = .012) events, and they selected Never for sad events more often than for happy events(p = .012). Thus, older children demonstrated a preference to select the Never box for frighteningevents more often than for the other emotions, suggesting that they might not know whether theseevents can occur.

In sum, the findings provide partial support for the study’s hypotheses. Children across agesreported that happy events could occur to themselves more often than frightening and sad events.There were, however, developmental differences in children’s judgments of what can occur to anotherchild or never. The 3-year-olds did not discriminate between which emotional events could occur toanother child and showed a bias based on emotional valence that frightening and sad events couldnever occur. The 4- and 5-year-olds supported the study’s hypotheses and generally selected thatsad events could occur to another child and frightening events could never occur.


Discussion

Study 2’s findings suggest that the sorting task was successful at teasing apart potential mecha-nisms underlying older children’s evaluations of reality. Specifically, do children’s judgments reflecttheir knowledge that an event can or cannot occur or, rather, are they a way of coping with the emo-tional information? Children’s preference to select Never for frightening images provides evidencethat children may lack the knowledge that these events can occur. Typically developing childrenare often sheltered from frightening events by their parents and, thus, might not have experience withor know that these events can occur. Results from Study 2 echo those of Carrick et al. (2010) that mal-treated children, who have routine exposure to negative events, were more likely to report that fright-ening events could occur than were nonmaltreated children. Of note, however, because evidencesuggests that children sometimes consider an improbable event as impossible (Shtulman & Carey,2007), an alternative explanation that cannot be ruled out is that children believed the frighteningevents were very unlikely to occur. Still, if this were the case, they should have selected the Otherbox for frightening events more often than they did.

Sad events, in contrast, appear to be evaluated differently from frightening events by olderchildren. When selecting which events could occur to another child, older children tended to selectsad events, indicating that they know sad events can occur but would prefer that they occur to some-one else. Feeling or witnessing sadness is a relatively normative experience during childhood; thus,children likely have created mental scripts that these events can occur. However, despite the possibil-ity of knowing that sad events can occur, children in Study 2 appeared to have employed a strategy tocope with the negative feelings of sadness by deeming that sad events can occur to others more oftenthan to themselves.

Of note, 3-year-olds’ selections for the Other and Never boxes were different from those of olderchildren. Their selection for the Other box was evenly split across emotions, suggesting that perhapsunderstanding what can occur to another child is too cognitively advanced for this age. They did, how-ever, show a response bias for the Never box that was based on the events’ valence (i.e., frighteningand sad events can never occur more often than happy events), consistent with 3-year-olds’ perfor-mance in Study 1. Thus, discrete emotions likely begin to influence children’s evaluations of fantasyand reality after 3 years of age.

The proposed interpretations of children’s box selections should be viewed with caution becausealternative interpretations are possible, particularly for the Other and Never boxes. Children may haveplaced in the Other box events that they deemed as more likely to occur to someone else, rather thanevents they would prefer occur to someone else, either because they possess an unrealistic belief thatbad things cannot occur to them or because they are generalizing from their experiences that theseevents do not occur to them. The Never box, which was interpreted as representing events childrendid not know could happen, may represent events that children do not want to occur or are unwillingto admit can occur. The alternative interpretations suggest that children have knowledge that sad andfrightening events can occur but use an emotion regulation strategy for frightening versus sad events.Further testing is necessary to better tease apart these interpretations.

General discussion

Studies 1 and 2 revealed important differences in how emotions and context influence children’sfantasy–reality distinctions. In general, the findings indicate that children show a bias to report thathappy and sad fantastic events can occur but that this bias can be reduced when the children arerewarded to responded accurately. Children’s evaluations of real events are also influenced by emo-tions, such that older children will report that happy and sad real events can occur but that they wouldprefer happy events to occur to themselves and sad events to occur to another child. Frighteningevents, on the other hand, are deemed as not able to occur even when children are rewarded to reportaccurately. It appears that children may understand the boundaries of fantasy but have some limits intheir understanding of reality. Several general conclusions can be made from these findings.

Both studies provide support for the unique contribution that discrete emotions have on children’sevaluations of fantasy and reality. The 4- and 5-year-olds showed distinct patterns of responses for


happy, sad, and frightening information across both fantasy and reality. This expands on prior studiesof fantasy–reality distinctions that either examined only general positive and negative information(Bourchier & Davis, 2000a, 2000b; Harris et al., 1991) or found differences in performance based solelyon valence (Carrick & Quas, 2006). The current findings are consistent with the growing literatureregarding the influence of discrete emotions on children’s cognition (Davis, Quas, & Levine, 2008;Levine, Burgess, & Laney, 2008; Trentacosta & Fine, 2010) and argue for the need to take any emotionalcontent into consideration when examining cognitive tasks.

The current findings also highlight that children actively construct their understanding of fantasticand real information. Regarding fantasy, Study 1’s finding that children were more likely to reportthat fantasy was not real when they were rewarded provides evidence that children are sensitiveto the context surrounding their evaluations of fantasy. They will likely engage in more fantasticthinking when the environment supports such thinking or does not directly restrict it. For example,research on fantastic thinking in Mennonite children, a culture that tends to discourage fantasy-ori-ented behaviors, found a discrepancy in children’s fantasy play when alone versus when in the pres-ence of disapproving adults (Carlson, Taylor, & Levin, 1998; Taylor & Carlson, 2000). Such aphenomenon is also seen in children’s general play behaviors; children will base their level of engage-ment in play on cues regarding whether or not the environment supports play (Rogers & Sawyers,1988; Rubin, Fein, & Vandenberg, 1983). Of note, the fact that children can easily move betweenendorsing and not endorsing fantasy attests to the likelihood that children are not generally credulousto all fantastic information and, in fact, use contextual information when judging the information’sreality status (Tullos & Woolley, 2009). When children show signs of belief in fantasy (e.g., SantaClaus, the Tooth Fairy), there is often considerable support from the environment, including adultencouragement.

Children’s beliefs about reality are also a result of piecing together information from their environ-ment. Here, unlike fantasy, however, children’s evaluations of reality appear to be less easily manip-ulated given the lack of improvement during Phase 2 of Study 1. In addition, children’s beliefs arelikely heavily based on personal experiences. Children routinely encounter happy and sad events,and so it is not surprising that they report these events can occur. However, Study 2 showed that theywould prefer happy events occur to themselves and sad events occur to another child, suggesting thattheir construction of reality is still influenced by personal desires. Frightening events, in contrast, arenot as routinely experienced or may even be explicitly labeled as not possible by parents (e.g., Fivushet al., 2003); thus, children do not see these events as viable. This raises the question of what children,who have been sheltered from negative events, will do if they actually experience a frightening event.Perhaps a developmentally appropriate discussion of what to do in these situations may be beneficial.Prior research has found individual differences in parent–child conversations about negative eventssuch as children’s gender, parent–child attachment style, and culture (Fivush, 1991; Sales, 2009).These factors are important to consider in future research to examine how individual differences inparental attitudes and behaviors regarding discussing negative events influence children’s evaluationsof such events.

The findings from Study 1 and 2 are somewhat consistent with an optimism bias seen in preschool-ers. That is, children reported that happy fantasy can occur and that happy, but not sad, reality canoccur to themselves while denying that frightening events can occur. Optimism biases have been seenin other areas of young children’s lives, including a heightening perception of their cognitive (e.g.,Stipek, 1981; Stipek & MacIver, 1989) and physical (e.g., Schneider, 1998; Schwebel & Plumert,1999) skills and a tendency to think positively about others with little evidence otherwise (Boseovski& Lee, 2006; see also Seligman, Reivich, Jaycox, & Gillham, 1995, for a general review). However, anoptimism bias does not completely explain why children reported that sad fantastic events can occur.Children should not necessarily want sad events to occur unless they were somehow drawn to the fan-tastic characters and felt empathy or concern for them. Additional research is necessary to test thisexplanation and the impact that optimism can have on young children’s fantasy–reality distinctions,particularly whether developmental changes in optimism coincide with changes in children’sdistinctions.

The studies have limitations that should be addressed in future research. First, as discussed, thereare alternative interpretations for the Other and Never boxes in Study 2, suggesting that children may


have used an emotion regulation strategy for frightening events rather than sad events. Second,although care was taken to ensure that the test images depicted the three emotions accurately, somechildren may have interpreted the target emotion inaccurately. Future research should include a moreprecise measure of children’s ability to identify the depicted emotion. Finally, the current studies didnot include measures of individual differences that could have influenced children’s performance.Often studies on fantasy–reality distinctions include measures of children’s proneness to engage infantasy-oriented activities in order to test whether higher fantasy orientation is associated with great-er or decreased performance (Sharon & Woolley, 2004; Woolley et al., 2004). However, given the cur-rent studies’ focus on emotion, including measures of emotion regulation and/or reactivity wouldlikely be more effective at uncovering important individual differences.

In conclusion, these findings show the importance of paying attention to the broader context ofhow children develop their beliefs about fantasy and reality and not assuming that children uniformlybelieve or dismiss the existence of fantastic information. Many factors, including emotional content,personal experiences, and testing contexts, contribute to the development of children’s beliefs andshould be considered when evaluating children’s understanding.

Acknowledgments

We thank Jodi Quas, Linda Levine, Thomas Lyon, Janna Kim, Angela Evans, Jennifer Piazza, KristyMatsuda, and Rachel Fenning for their suggestions on the manuscript. We thank Shelley Manzer, EliseSchaffer, Destiny Osmialowski, Kristen Schaffer, Tiffany Bui, and Alex Harris for their assistance incollecting and coding the data, and we thank Leigha Panelli for drawing the images.

Appendix A

Descriptions of Study 1 test images.

Emotion
Fantastic events Real events
Happy
Laughing baby panda in a stroller Mother and son hugging Dancing mice wearing clothing Mother smiling holding baby Raccoons hugging Grandmother laughing with girl Two laughing mice dancing Father picking up son and smiling Happy giant laughing with people Boy smiling with mother Mice smiling and sitting on a sandwich Family hugging
Frightening
Frightening giant House on fire Fire-breathing dragon Two scary snakes Giant scaring children Group of frightening wolves Goblin scaring a boy Wolf chasing child Group of frightening witches Dog biting a boy Flying dragon hurting a man Cats fighting
Sad Bear dressed in pajamas crying Mother crying with children
Pig dressed in a suit crying Husband with wife crying Bear crying in a house Girl crying because lost Gerbil crying because dropped ice cream cone Woman crying Bunny crying in a field Father hugging crying girl Ram crying because broke a dish Woman consoling crying girl


Appendix B

Descriptions of Study 2 test images.

Theme
Emotion Vignette
Child and house
Happy This child is so happy to be playing in front of house Frightening This child is so scared because the house is on fire Sad This child is so sad because she lost something and has to
find it

Child and parentbaking

Happy
This child and parent are so happy to be baking cookies
Frightening
This child and parent are so scared because oven is on fire Sad This child and parent are so sad because cookies are burnt
Child and dog
Happy This child is so happy to be eating ice cream with her dog Frightening This child is so scared because the dog is trying to get the ice
cream
Sad This child is crying because the ice cream has fallen next to
the dog

Child and nest
Happy This child is so happy to have found a nest of bird eggs Frightening This child is so scared because a snake is trying to eat the
bird eggs
Sad This child is so sad because a bird egg is broken
References

Berkowitz, L. (1999). Anger. In T. Dalgleish & M. J. Power (Eds.), Handbook of cognition and emotion (pp. 411–428). New York:John Wiley.

Boseovski, J. J., & Lee, K. (2006). Children’s use of frequency information for trait categorization and behavioral prediction.Developmental Psychology, 42, 500–513.

Bourchier, A., & Davis, A. (2000a). Individual and developmental differences in children’s understanding of the fantasy–realitydistinction. British Journal of Developmental Psychology, 18, 353–368.

Bourchier, A., & Davis, A. (2000b). The influence of availability and affect on children’s pretence. British Journal of DevelopmentalPsychology, 18, 137–156.

Browne, C. A., & Woolley, J. D. (2004). Preschoolers’ magical explanations for violations of physical, social, and mental laws.Journal of Cognition and Development, 5, 239–260.

Carlson, S., Taylor, M., & Levin, G. (1998). The influence of culture on pretend play: The case of Mennonite children. Merrill-Palmer Quarterly, 44, 538–565.

Carrick, N., & Quas, J. A. (2006). The effects of discrete emotions on young children’s ability to discern fantasy and reality.Developmental Psychology, 42, 1278–1288.

Carrick, N., Quas, J. A., & Lyon, T. D. (2010). Maltreated and nonmaltreated children’s evaluations of emotional fantasy. ChildAbuse & Neglect, 4, 129–134.

Carver, C., & White, T. (1994). Behavioral inhibition, behavioral activation, and affective responses to impending reward andpunishment: The BIS/BAS scales. Journal of Personality and Social Psychology, 67, 319–333.

Cook, C., & Sobel, D. M. (2011). Children’s beliefs about the fantasy/reality status of hypothesized machines. DevelopmentalScience, 14, 1–8.

Davidson, R. J. (1998). Affective style and affective disorders: Perspectives from affective neuroscience. Cognition & Emotion, 12,307–330.

Davis, E., Quas, J. A., & Levine, L. (2008). Children’s memory for stressful events: Exploring the role of discrete emotions. In M.Howe, G. Goodman, & D. Cicchetti (Eds.), Stress, trauma, and children’s memory development: Neurobiological, cognitive,clinical, and legal perspectives (pp. 236–264). New York: Oxford University Press.

Eisenberg, N., & Fabes, R. (1992). Emotion, regulation, and the development of social competence. In M. S. Clark (Ed.), Emotionand social behavior (pp. 119–150). Thousand Oaks, CA: Sage.

Eisenberg, N., Fabes, R., Murphy, B., Karbon, M., Maszk, P., Smith, M., et al (1994). The relations of emotionality and regulation todispositional and situational empathy-related responding. Journal of Personality and Social Psychology, 66, 776–797.

Fabes, R., Eisenberg, N., McCormick, S., & Wilson, M. (1988). Preschoolers’ attributions of the situational determinants of others’naturally occurring emotions. Developmental Psychology, 24, 376–385.


Fivush, R. (1991). Gender and emotion in mother–child conversations about the past. Journal of Narrative & Life History, 1,325–341.

Fivush, R., Berlin, L. J., Sales, J. M., Mennuti-Washburn, J., & Cassidy, J. (2003). Functions of parent–child reminiscing aboutemotionally negative events. Memory, 11, 179–192.

Gray, J. A. (1981). A critique of Eysenck’s theory of personality. In H. J. Eysenck (Ed.), A model for personality (pp. 246–276).Berlin: Springer-Verlag.

Gray, J. A. (1990). Brain systems that mediate both emotion and cognition. Cognition & Emotion, 4, 269–288.Guthrie, I., Eisenberg, N., Fabes, R., Murphy, B., Holmgren, R., Mazsk, P., et al (1997). The relations of regulation and emotionality

to children’s situational empathy-related responding. Motivation and Emotion, 21, 87–108.Harmon-Jones, E., & Allen, J. (1998). Anger and frontal brain activity: EEG asymmetry consistent with approach motivation

despite negative affective valence. Journal of Personality and Social Psychology, 74, 1310–1316.Harmon-Jones, E., & Sigelman, J. (2001). State anger and prefrontal brain activity: Evidence that insult-related relative left-

prefrontal activation is associated with experienced anger and aggression. Journal of Personality & Social Psychology, 80,797–803.

Harris, P., Brown, E., Marriott, C., Whittall, S., & Harmer, S. (1991). Monsters, ghosts, and witches: Testing the limits of thefantasy–reality distinction in young children. British Journal of Developmental Psychology, 9, 105–123.

Hoffman, M. L. (1985). Interaction of affect and cognition in empathy. In C. Izard, J. Kagan, & R. Zajonc (Eds.), Emotions, cognition,and behavior (pp. 103–131). New York: Cambridge University Press.

Lazarus, R. S. (1991). Emotion and adaptation. Oxford, UK: Oxford University Press.Levine, L. (1995). Young children’s understanding of the causes of anger and sadness. Child Development, 66, 697–709.Levine, L., Burgess, S., & Laney, C. (2008). Effects of discrete emotions on young children’s suggestibility. Developmental

Psychology, 44, 681–694.Quas, J. A., Wallin, A. R., Horwitz, B., Davis, E., & Lyon, T. D. (2009). Maltreated children’s understanding of and emotional

reactions to dependency court involvement. Behavioral Sciences and the Law, 27, 97–117.Rogers, C. S., & Sawyers, J. K. (1988). Play in the lives of children. Washington, DC: National Association for the Education of Young

Children.Roseman, I. (1991). Appraisal determinants of discrete emotions. Cognition & Emotion, 5, 161–2000.Rubin, K. H., Fein, G. G., & Vandenberg, B. (1983). Play. In P. H. Mussen (Ed.). Handbook of child psychology (Vol. 4, pp. 693–774).

New York: John Wiley.Sales, J. (2009). Creating a context for children’s memory: The importance of parental attachment status, coping, and narrative

skill for co-constructing meaning following stressful experiences. In J. A. Quas & R. Fivush (Eds.), Emotion and memory indevelopment: Biological, cognitive, and social considerations (pp. 196–217). New York: Oxford University Press.

Samuels, A., & Taylor, M. (1994). Children’s ability to distinguish fantasy events from real-life events. British Journal ofDevelopmental Psychology, 12, 417–427.

Sayfan, L., & Lagattuta, K. (2009). Scaring the monster away: What children know about managing fears of real and imaginarycreatures. Child Development, 79, 821–835.

Schneider, W. (1998). Performance prediction in young children: Effects of skill, metacognition, and wishful thinking.Developmental Science, 1, 291–297.

Schwebel, D. C., & Plumert, J. M. (1999). Longitudinal and concurrent relations among temperament, ability estimation, andinjury proneness. Child Development, 70, 700–712.

Seligman, M. E. P., Reivich, K., Jaycox, L., & Gillham, J. (1995). The optimistic child. Boston: Houghton Mifflin.Sharon, T., & Woolley, J. D. (2004). Do monsters dream? Young children’s understanding of the fantasy/reality distinction. British

Journal of Developmental Psychology, 22, 293–310.Shtulman, A., & Carey, S. (2007). Improbable or impossible? How children reason about the possibility of extraordinary events.

Child Development, 78, 1015–1032.Skolnick, D., & Bloom, P. (2006). What does Batman think about SpongeBob? Children’s understanding of the fantasy/fantasy

distinction. Cognition, 101, B9–B18.Skolnick, D., & Bloom, P. (2009). Young children separate multiple pretend worlds. Developmental Science, 12, 699–705.Stein, N., & Levine, L. J. (1999). The early emergence of emotional understanding and appraisal: Implications for theories of

development. In T. Dalgleish & M. Power (Eds.), Handbook of cognition and emotion (pp. 383–408). New York: John Wiley.Stipek, D. J. (1981). Children’s perceptions of their own and their classmates’ ability. Journal of Educational Psychology, 73,

404–410.Stipek, D. J., & MacIver, D. (1989). Developmental change in children’s assessment of intellectual competence. Child

Development, 60, 521–538.Subbotsky, E. (1994). Early rationality and magical thinking in preschoolers: Space and time. British Journal of Developmental

Psychology, 12, 97–108.Subbotsky, E. (2001). Causal explanations of events by children and adults: Can alternative causal modes coexist in one mind?

British Journal of Developmental Psychology, 19, 23–45.Subbotsky, E., & Slater, E. (2011). Children’s discrimination of fantastic vs. realistic visual displays after watching a film with

magical content. Perceptual and Motor Skills, 112, 603–609.Taylor, M., & Carlson, S. (2000). Influence of religious beliefs on parental attitudes. In K. Rosengren, C. Johnson, & P. Harris (Eds.),

Imagining the impossible (pp. 247–268). Cambridge, UK: Cambridge University Press.Taylor, M., Shawber, A. B., & Mannering, A. M. (2009). Children’s imaginary companions: What is it like to have an invisible

friend? In K. Markman, W. Klein, & J. Suhr (Eds.), Handbook of imagination and mental simulation (pp. 211–224). New York:Psychology Press.

Trentacosta, C. J., & Fine, S. E. (2010). Emotion knowledge, social competence, and behavior problems in childhood andadolescence. A meta-analytic review. Social Development, 19, 1–29.

Tullos, A., & Woolley, J. D. (2009). The development of children’s ability to use evidence to infer reality status. Child Development,80, 101–114.


Wacker, J., Heldmann, M., & Stemmler, G. (2003). Separating emotion and motivational direction in fear and anger: Effects offrontal asymmetry. Emotion, 3, 167–193.

Widen, S., & Russell, J. A. (2010). Differentiation in preschoolers’ categories of emotion. Emotion, 10, 651–661.Woolley, J., Boerger, E. A., & Markman, A. B. (2004). A visit from the Candy Witch: Factors influencing young children’s belief in a

novel fantastical being. Developmental Science, 7, 456–468.Woolley, J., & Cox, V. (2007). Development of beliefs about storybook reality. Developmental Science, 10, 681–693.Woolley, J., & Phelps, K. E. (1994). Young children’s practical reasoning about imagination. British Journal of Developmental

Psychology, 12, 53–67.Zahn-Waxler, C., Radke-Yarrow, M., Wagner, E., & Chapman, M. (1992). Development of concern for others. Developmental

Psychology, 28, 126–136.

Documents

Preschoolers’ fantasy–reality distinctions of emotional events