Questioning as a component of scaffolding in predicting emotion knowledge in preschoolers

This article was downloaded by: [George Mason University]On: 06 June 2013, At: 12:23Publisher: RoutledgeInforma Ltd Registered in England and Wales Registered Number: 1072954 Registeredoffice: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK

Early Child Development and CarePublication details, including instructions for authors andsubscription information:http://www.tandfonline.com/loi/gecd20

Questioning as a component ofscaffolding in predicting emotionknowledge in preschoolersCraig S. Bailey a , Susanne A. Denham a & Timothy W. Curby aa Psychology Department, George Mason University, 4400University Drive, MS 3F5, Fairfax, VA, 22030, USAPublished online: 16 Apr 2012.

To cite this article: Craig S. Bailey , Susanne A. Denham & Timothy W. Curby (2013): Questioningas a component of scaffolding in predicting emotion knowledge in preschoolers, Early ChildDevelopment and Care, 183:2, 265-279

To link to this article: http://dx.doi.org/10.1080/03004430.2012.671815

PLEASE SCROLL DOWN FOR ARTICLE

Full terms and conditions of use: http://www.tandfonline.com/page/terms-and-conditions

This article may be used for research, teaching, and private study purposes. Anysubstantial or systematic reproduction, redistribution, reselling, loan, sub-licensing,systematic supply, or distribution in any form to anyone is expressly forbidden.

The publisher does not give any warranty express or implied or make any representationthat the contents will be complete or accurate or up to date. The accuracy of anyinstructions, formulae, and drug doses should be independently verified with primarysources. The publisher shall not be liable for any loss, actions, claims, proceedings,demand, or costs or damages whatsoever or howsoever caused arising directly orindirectly in connection with or arising out of the use of this material.

Questioning as a component of scaffolding in predicting emotionknowledge in preschoolers

Craig S. Bailey∗, Susanne A. Denham and Timothy W. Curby

Psychology Department, George Mason University, 4400 University Drive, MS 3F5,Fairfax, VA 22030, USA

(Received 15 January 2012; final version received 28 February 2012)

The following study expands Denham and Auerbach’s (1995. Mother-childdialogue about emotions and preschoolers’ emotional competence. Genetic,Social, and General Psychology Monographs, 121, 313–337) findings,demonstrating a link between mothers’ talk about emotions and preschoolers’knowledge of emotions. We investigate the maternal language and child emotionknowledge association by analysing all maternal questioning, not just questionsspecifically about emotions, during a mother–child interaction task. Questionsare functionally different from other features of language because questions pushchildren to think and formulate their own thoughts about the content, in our case,emotions, in order to generate an appropriate response. Twenty-eight transcriptsof mother–child dialogue during a wordless storybook task with emotionallysalient content were coded for mothers’ use of questions. Results show thatmothers’ use of questions was associated with emotion knowledge, but only forgirls. Our findings show how emotions are socialised differently for boy andgirls and the potential importance that questions play in scaffolding girls’attainment of emotion knowledge and overall emotional competence.

Keywords: questioning; scaffolding; emotion knowledge; gender differences;early childhood; emotion socialisation

Research has documented that the language used by parents is predictive of children’sunderstanding of emotions and their subsequent competence in relating with others(Cervantes & Callanan, 1998; Denham & Auerbach, 1995; Denham, Cook, &Zoller, 1992; Garner, Carlson Jones, Gaddy, & Rennie, 1997; Hart, Newell, &Olsen, 2003; Warren & Stifter, 2008). Specifically, mothers who use more emotionwords to describe emotionally important information have children who demonstratemore knowledge of emotions and emotionally salient situations (Denham & Auerbach,1995). Use of emotion words during conversation, however, may not be the onlymeans by which adults socialise children’s developing knowledge of emotions.During dialogue, mothers use a range of linguistic devices to convey meaning andbuild joint understanding about emotions. Unlike statements, questions are structurallyand functionally different in ways that promote learning (Blewitt, Rump, Shealy, &Cook, 2009). Questions, such as, ‘How does the puppy feel?’ place a higher cognitivedemand on the listener than statements like, ‘The puppy feels sad’. Mothers’ use ofquestions may therefore be serving to scaffold children’s development of emotion

# 2013 Taylor & Francis

∗Corresponding author. Email: [email protected]

Early Child Development and Care, 2013Vol. 183, No. 2, 265–279, http://dx.doi.org/10.1080/03004430.2012.671815

Dow

nloa

ded

by [

Geo

rge

Mas

on U

nive

rsity

] at

12:

23 0

6 Ju

ne 2

013

knowledge. In the current study, we investigate whether mothers who ask morequestions of their preschool-aged children during a wordless book-reading task havechildren who are more knowledgeable of emotions.

Emotion knowledge

Children who are more knowledgeable of emotions are more prepared for the socialand emotional demands of school (Denham, Zinsser, & Brown, in press), are moreliked by teachers and peers (Denham, McKinley, Couchoud, & Holt, 1990), and aremore pro-social in situations where adults display anger, sadness, or pain (Denham& Couchoud, 1991). The current investigation focuses on emotion knowledge asone aspect of emotional competence (Denham, 2007). Emotional competenceitself includes not only emotion knowledge, but also emotion regulation andemotion expression. Emotion knowledge reflects children’s abilities to correctlyidentify the emotions of others and themselves. The development of emotion knowl-edge is critical as many social interactions rely on the ability to correctly identifythe emotions being expressed (Denham, 2007; Halberstadt, Denham, & Dunsmore,2001; Saarni, Mumme, & Campos, 1998). In this way, development of emotionexpression and regulation are in part contingent on the development of emotionknowledge. Furthermore, parents work to help children develop their understand-ings of emotions (Eisenberg, Cumberland, & Spinrad, 1998). The present studyexamines whether asking questions is an effective way for parents to socialisechildren’s emotion knowledge.

Socialisation of emotion knowledge

Because emotions are an integral component of social interaction, success or failure ininteracting is due, in part, to the emotional processes between children and those aroundthem (Denham, 2007; Halberstadt et al., 2001; Saarni et al., 1998). Important adults likeparents do much to socialise children’s emotional competence (Eisenberg et al., 1998),explaining partially why some children are more emotionally competent than otherchildren. In order to understand fully how emotion knowledge develops, researchersmust investigate how features of parent–child interactions influence children’sacquisition of emotion knowledge.

One of the ways to understand how parents teach children emotion knowledge is toinvestigate how children learn about emotions through language. Children learn to useemotion-related words from their parents (see Cervantes & Callanan, 1998; Denham &Auerbach, 1995), giving them the ability to understand and to express emotions usinglanguage. Furthermore, emotion words provide children with insight and the necessaryvocabulary to understand and explain the antecedents and consequences of emotion.Exploring how parents socialise children’s emotion knowledge through dialogue istherefore of considerable merit.

The process by which language informs children about emotions during parent–childinteractions is twofold. First, language carries emotional content conveying informationabout how a parent feels through features of the language, for example, pitch and tone.Second, parents use emotion words to implicitly and explicitly discuss emotions andteach emotion knowledge. Specifically, parents use emotion words to qualify, comment,explain, guide, socialise, and question emotional content (Denham & Auerbach, 1995).When qualifying, a parent will add meaning to an event, e.g. ‘They’re real surprised

266 C.S. Bailey et al.

Dow

nloa

ded

by [

Geo

rge

Mas

on U

nive

rsity

] at

12:

23 0

6 Ju

ne 2

013

when they see the box’ (p. 335). When commenting, a parent will describe an event, e.g.‘The little boy is laughing and saying “ha”’ (p. 335). When explaining, a parent will inter-pret an event, e.g. ‘The boy went and got the dog in his arms and [because of this] every-body’s happy again’ (p. 335). When guiding, a parent will direct attention to behaviouralaspects of an event, e.g. ‘They’ve got the puppy so he won’t get hurt by the school bus’ (p.335). When socialising, parents will use an event as an explicit exemplar, e.g. ‘You shouldnever scold, [and] never hit, a dog’ (p. 335). When questioning, the focus of the presentstudy, a parent will request information or a response about an event, e.g. ‘How do youthink she feels?’ (p. 335) The importance of the distinctions in how a parent uses the fea-tures of language in dialogues with their children is that some language features, such asquestions, may carry more weight in teaching their children about emotions than theothers, or may promote thinking about emotions in a different way.

Questions?

Vygotsky placed great emphasis in his sociocultural conceptualisation of develop-ment on the role of language (Vygotsky, 1978). He believed that children learnwithin the interactions they have with adults, based in part on the language thatadults provide. Although it could be argued that all language is important for chil-dren’s development, questions are special in that they are functionally differentfrom other features of language because questions push children to think and formu-late their own thoughts about content in order to generate an appropriate response(Honig & Wittmer, 1981). Even rhetorical questions have an element of ambiguitydifferent from other features of language. Questions therefore serve as a componentof the scaffolding parents use when teaching or interacting with their children becausequestions allow parents to assess and match children’s Zone of Proximal Development(Vygotsky, 1978; Wood, Bruner, & Ross, 1976). In other words, questions are a toolthat parents can use to quickly assess children’s understanding of emotion and tomodify their responses accordingly.

Questions also serve the purpose of activating cognitive processes. Educationalresearch suggests that teachers utilise questions to engage their students (Adams, 1994),to promote critical thinking (Adams, 1994; Carr, 1998), and to improve comprehension(Weinstein, McDermott, & Roediger, 2010). In one study, scaffolding in the form ofasking questions about target words was found to improve three-year-old children’s voca-bulary (Blewitt et al., 2009). The questions proposed by teachers and parents to childrencan ignite curiosity and the self-generation of questioning that are foundational to latersuccess (Goodman & Berntson, 2000) and cognitive development (Chouinard, 2007).

Parents who scaffold by using questions have more positive parent–child inter-actions and have infants who are more emotionally engaged in their interactions(Barr, Zack, Garcia, & Muentener, 2008). The use of questions as a teaching toolimplicates questioning as an element of emotion socialisation. In fact, 61% of thequestions teachers ask toddlers contain wishes, desires, or feelings (Honig &Wittmer, 1981), demonstrating not only that the discussion of emotions with childrenoften involves the use of questions but also that questions are a component of emotionsocialisation.

Questions and the development of emotion knowledge

A wordless storybook reading task is used in the present study. In this task, the motheruses the pictures in an emotionally laden, wordless picture book to tell a story to the

Early Child Development and Care 267

Dow

nloa

ded

by [

Geo

rge

Mas

on U

nive

rsity

] at

12:

23 0

6 Ju

ne 2

013

child. During the task, the child listens to the mother’s narration and connects the wordsand ideas to the illustrations in the book. In this situation, the mother could elaboratevery little and not ask the child any questions, or the mother could generate commentsand ask questions about the character’s emotions, what caused those emotions, and theconsequences of those emotions based on the events of the story, thereby expanding theinformation outside the story and engaging the child in thinking about emotions.

This task highlights four features of parent–child interactions that are important inconsidering questions as part of the socialisation of emotions. First, emotions can be feltand seen, but explaining, labelling, and understanding emotions in others and the selfare primarily done through language. Second, parents differ in their reading strategiesand how they discuss emotions and, thus, how they socialise emotions. Third, question-ing by parents assists in scaffolding children’s acquisition of emotion knowledgeduring the storybook task. Fourth, parents’ use of questions about emotions andemotion situations challenges children’s thinking about emotions and expands theirknowledge of emotions, especially in contexts ripe for learning, like reading anemotionally laden storybook – a common occurrence in early childhood.

What remains to be answered is whether having children think about contexts andsituations in a story containing emotions could potentially lead children to acquire moreknowledge of emotions and emotion situations with parental assistance. Interpretationof emotions embedded in social information begins with the encoding of social cues,for example, vocal, facial, or behavioural displays of emotion (Halberstadt et al.,2001). By strategically using questions, parents are able to draw attention to thesalient cues of the emotional world. Once joint attention has been established, languageplays an even larger part in the development of emotion knowledge (Denham &Auerbach, 1995) because emotion words provide labels onto which the social–emotional cues can be mapped. For example, ‘Is the boy sad?’ not only draws attentionto the boy exhibiting behavioural features of sadness, but it also provides the label‘sad’, which identifies these behaviours. The question implies ambiguity, whichprompts a response by the child. Using questions when reading storybooks engagesthe child and turns the story into a thought-provoking exchange instead of a passivelistening experience.

In complex situations, there are often conflicting cues, making interpretationunlcear. Especially for young children, questions serve to scaffold the understandingof complicated social–emotional situations and further the development of emotionknowledge. To elaborate, questions point children in the right direction by highlightingcues that can lead children to the correct interpretation. Carefully crafted questionsprompt a response from the children that pushes their understanding of social–emotional information beyond their current level of understanding.

The role of children’s gender

The effect of mothers asking questions in situations with emotionally important infor-mation on acquisition of emotion knowledge may not be the same for boys as it is forgirls. Gender differences are born primarily through socialisation from caregivers andare embedded within culture (Kennedy Root & Denham, 2010; Leaper, 2002). Therules of gender-typed behaviour that young children acquire are the result of parents’beliefs and attitudes about gender rules and encouragement of gendered behaviour(Kennedy Root & Denham, 2010; Lytton & Romney, 1991; Tenenbaum & Leaper,2002). A meta-analysis by Leaper, Anderson, and Sanders (1998) found that mothers

268 C.S. Bailey et al.

Dow

nloa

ded

by [

Geo

rge

Mas

on U

nive

rsity

] at

12:

23 0

6 Ju

ne 2

013

use more supportive language with daughters than with sons (Leaper et al., 1998).Mother–daughter and mother–son socialisation may have different processes atplay. Additionally, Garner et al. (1997) found that daughters receive more modellingof positive emotions than sons and sons receive more modelling of anger thandaughters.

Discussions of emotions

Kennedy Root and Denham (2010) describe modelling of emotions, along with theoverall emotional climate of the family, as indirect processes of socialising emotions.Contingent reactions to emotions, and teaching of emotions and discussions ofemotions and emotion situations, are viewed as a more direct pathway in the socialisa-tion of emotion (Kennedy Root & Denham, 2010). These more direct pathways,especially discussion of emotions, offer a glimpse at how children come to acquire agendered understanding of emotions. For example, mothers go into greater detail andplace more stress on the interpersonal nature of emotions with daughters than sons(Fivush, Berlin, Sales, Mennuti-Washburn, & Cassidy, 2003; Kuebli, Butler, &Fivush, 1995). These findings indicate that any effect of gender might even be exacer-bated when situations involve interpersonal contexts, such as when discussing the inter-actions of characters in storybooks.

It is less clear whether parents, in contexts that are heavily interpersonal and thusprimed for discussions of emotions, would ask more questions of a particular gender.Cervantes and Callanan (1998) investigated gender differences in mother–child talkabout emotions when children were two, three, and four years old. The authors opera-tionalised explanations of emotions in a way that also coded for mothers’ questions, aslong as those questions elicited a response from the child. Results indicated thatmothers used more explanations of emotions than labels of emotion in their talk toboys, after controlling for age. There was no difference in the explanation of emotionsversus labels of emotions in mothers’ emotion talk for girls (Cervantes & Callanan,1998). These results point to a pattern where mothers may ask more questions ofboys than girls in contexts involving emotions.

However, Fivush et al. (2003) found that mothers explain more about emotions togirls than boys and this likewise extends to findings that show that caregivers, ingeneral, ask more questions of girls than boys (Cherry & Lewis, 1976; Clearfield &Nelson, 2006). Furthermore, Denham and Auerbach (1995) found no gender differencesin the number of maternal emotion utterances that were questions, which corroboratesHonig and Wittmer’s (1981) results that show no gender differences in the number ofquestions about feelings, wishes, and desires of self and others. In investigations thatare not particular to emotion, others have found similar findings of no difference ofgender in questions asked (Kruper & Uzgiris, 1987). These conflicting findings lead tothe current inclusion of child gender in the investigation of whether mothers’ use ofquestions during a storybook is not only important for children’s developing knowledgeof emotions, but also whether mothers ask more questions of a particular gender and ifthis gender difference matters in the acquisition of emotion knowledge.

The current study

Denham and Auerbach’s (1995) demonstrated a link between mothers’ talk aboutemotions and preschoolers’ knowledge of emotions through an emotionally evocative,

Early Child Development and Care 269

Dow

nloa

ded

by [

Geo

rge

Mas

on U

nive

rsity

] at

12:

23 0

6 Ju

ne 2

013

wordless storybook task. Although Denham and Auerbach included maternal question-ing, they operationalised maternal questions as those uttered by mothers strictly invol-ving emotion words. We take a different approach to questions and build on their studyby including those questions that may very well be about emotions but do not includeemotion words. Particularly in the context of a storybook designed to be aboutemotions, questions about the characters and the plot may still help children becomemore knowledgeable about emotions. We hypothesise that maternal use of questioningscaffolds their development of emotion knowledge, as demonstrated by the task. Inaddition, we believe that this effect may be different for boys and girls based in parton the nature of the storybook. For these reasons, we seek to answer (a) whethermothers ask more questions of boys or girls during an emotionally charged storybookreading, (b) whether mothers who ask more questions have children who are moreknowledgeable about emotions, and (c) whether asking questions is particularlysalient for boys or girls learning emotion knowledge.

Method

Participants

Mother–child dyads were part of a larger longitudinal study investigatingsocial–emotional development during early childhood (see Denham & Auerbach,1995) and included 32 mother–child dyads with complete data on all variables outof a possible 71 dyads. Overall, all families were of middle- to upper-middle socioeco-nomic status. Half the children were boys, and missing data analyses showed that thatthere were no more boys than girls with missing data on the affective knowledgetest (AKT, explained below), x2 (1) ¼ 1.50, p ¼ 0.22 or storybook transcripts, x2

(1) ¼ 0.13, p ¼ 0.72. Children ranged in age from 30 to 54 months (M ¼ 41.11,SD ¼ 6.11).

Measures

Emotion knowledge

To assess the children’s emotion knowledge, the AKT (Denham & Auerbach, 1995)was used. The AKT uses puppets with removable faces indicating four emotions (i.e.happy, sad, angry, afraid). Administering the AKT proceeds through a series ofsteps. First, when presented in random order, the child is asked to name each facialexpression. Second, the experimenter asks the child to identify a particular emotionusing the faces. Third, the puppets are used to depict 20 vignettes where the puppeteeracts out 20 different situations with the puppets while using affective vocal and visualcues. Vignettes are age appropriate, such as receiving a gift to show happiness orhaving a nightmare to indicate fear. After each vignette, children are asked to indicatethe depicted emotion. Eight of the vignettes are designed to assess a child’s stereo-typical emotion knowledge by having the puppet act out situations with emotionsthat most children would feel (e.g. being happy about getting an ice cream cone).The remaining 12 vignettes are designed to assess a child’s non-stereotypical emotionknowledge by having the puppet act out situations with emotions that are differentfrom what the child’s mother had reported in a questionnaire (e.g. if the mother hadreported that her child was happy to go to school, the puppet would act out fear orsadness). Of these 12 vignettes, 6 involve positive situations with negative emotionswhereas the other 6 involve negative situations with another negative emotion.

270 C.S. Bailey et al.

Dow

nloa

ded

by [

Geo

rge

Mas

on U

nive

rsity

] at

12:

23 0

6 Ju

ne 2

013

Two points were awarded for correct identification of emotions and only one pointfor identifying the correct valence but the wrong emotion for each section. Points areaggregated for each component of the AKT into a composite score and used as theoverall score of emotion knowledge (a ¼ 0.91).

Questioning

Mother–child speech was observed during an emotionally laden, wordless storybookexperience. During a lab visit, mothers were given storybooks and were asked tonarrate the pictures to their children. The storybook was 23 pages, depicting a familythat receives a new puppy, with each page involving a scene designed to elicitvarious emotions, such as happiness, surprise, sadness, anger, and fear. The mother–child interaction was videotaped and transcribed; utterances were coded from the tran-scripts. An utterance was defined as a section of transcribed speech between pauses;usually sentences (Winsler, Fernyhough, McClaren, & Way, 2005). If an utterancestood alone as a separate clause, the sentence contained two, rather than one, utterances.For example, ‘the rabbit bit the boy, didn’t he?’ would be considered two total utter-ances with one utterance as a question because ‘the rabbit bit the boy’ is a separate utter-ance from ‘didn’t he?’ ‘Didn’t he?’ is eliciting a response to answer the question ‘didthe rabbit bite the boy?’ regardless of whether the child produces an audible response.

The authors counted the total number of utterances and the number of those utter-ances that were questions, regardless of whether questions were specifically aboutemotions or included emotional vocabulary (c.f. Denham & Auerbach, 1995).

Proportion of questioning. In order to account for mothers who were simply moreverbally expressive, the proportion of utterances that were questions was calculatedby taking the utterances classified as questions and dividing by the total utterances(Kruper & Uzgiris, 1987).

Results

Preliminary data analysis

Assumptions

All variables, their residuals, and their combined Mahalanobis distances were inspectedfor outliers and assessed for univariate and multivariate normality using boxplots,scatterplots, histograms, and Q–Q plots. Four outliers were identified beyond 2.5 stan-dard deviations and removed from further analyses. This procedure was primarilyundertaken to ensure that children with very low or very high levels of emotion knowl-edge and mothers with very high or low rates of questioning did not distort analyses. Allvariables were normally distributed after removal of outliers (see Table 1 for skewnessand kurtosis coefficients).

Descriptives

Descriptive statistics found in Table 1 show the averages for both genders on the AKT,Total Maternal Utterances, Total Maternal Utterances as Questions, Proportion of Ques-tions to Utterances, and Age. Inspecting minimum and maximum values found that,overall, some mothers were very talkative, using 343 utterances during the 23-pagestory whereas other mothers were not, only using 54 utterances. Similarly, the same

Early Child Development and Care 271

Dow

nloa

ded

by [

Geo

rge

Mas

on U

nive

rsity

] at

12:

23 0

6 Ju

ne 2

013

was true for mothers’ use of questions, ranging from 12 to 143 questions. Also found inTable 1 are results from t-tests revealing that boys were not found to be older than girls.However, significant gender differences were found with respect to the Proportion ofQuestioning, with mothers asking proportionally more questions of their daughters thansons. No significant differences between genders were found with respect to the AKT.

Primary data analysis

A three-step hierarchical regression was used to (a) control for age, (b) test whethermothers who ask more questions during interactions with their children have childrenwho are more emotionally competent, and (c) test whether child gender moderates theassociation between questioning and emotional competence. Age of child was used as acontrol in the first step of the regression, so that the other predictors’ contributionscould be evaluated over and above Age with Proportion of Questions and Gender inthe next step, and the Gender-Proportion of Questions interaction term in the finalstep. Results are found in Table 2.

All variables entered in the hierarchical regression were centred to avoid multicol-linearity. Centring also allows for the direct interpretation of betas (Cohen, Cohen,West, & Aiken, 2002). The interaction term used to test for moderation was madefrom the centred Gender and Proportion of Questioning variables.

According to Step 1, Age by itself contributes significantly to the prediction ofemotion knowledge via the AKT. Gender and Proportion of Questions neitherpredicted a significant proportion of the variance over and above Age when enteredin the second step, △R2 ¼ 0.10 (adjusted △R2 ¼ 0.04), △F (2, 24) ¼ 2.55, p ¼0.08, nor were their coefficients significant. When the interaction between Proportionof Questions and Gender was included in the model in the final step, a significantamount of variance was explained, but only marginally over and above Age, Gender,and Proportion of Questioning, △R2 ¼ 0.10 (adjusted △R2 ¼ 0.09), △F (1, 23) ¼4.11, p ¼ 0.06. The Gender X Proportion of Questions Interaction thus approachedtraditional significance in predicting emotion knowledge scores on the AKT. Secondary

Table 1. Descriptive statistics and significant differences between boys and girls on the AKT.

AKTaTotal maternal

utterancesTotal maternal

questions

Proportion ofquestions toutterancesb Agec

N 47d 41 41 41 71

Total 3.31 (6.90) 157.75 (64.24) 48.71 (25.88) 0.31 (0.08) 41.11 (6.11)

Boys 1.92 (7.89) 161.50 (51.82) 44.50 (17.89) 0.28 (0.08) 42.36 (6.87)

Girls 4.70 (5.70) 154.00 (76.52) 52.93 (32.14) 0.34 (0.06) 39.86 (5.19)

Skewness 21.16 0.96 1.91 20.60 0.39

Kurtosis 1.12 1.10 5.42 0.40 20.55

t-Test 21.07 0.30 20.86 22.15∗ 1.09

aAffective knowledge test.bTotal questions divided by total utterances.cAge in months.dAfter outliers removed. All variables are reported prior to centring, but after removing outliers∗p , 0.05.

272 C.S. Bailey et al.

Dow

nloa

ded

by [

Geo

rge

Mas

on U

nive

rsity

] at

12:

23 0

6 Ju

ne 2

013

analyses, which take into account missing data, found this term to be significant. Findingstatistical significance for interaction terms is especially difficult after the variance formain effects are removed (Cohen et al., 2002). With our sample size of 28 and themedium effect size of f 2 ¼ 0.17, which is based on the semi-partial correlation forthe interaction term and on a total R2 ¼ 0.35, we feel that interpreting coefficientsapproaching traditional significant levels is warranted (Cohen, 1992).

As shown in Figure 1, the association between mothers’ use of questions and theirchildren’s knowledge of emotions was different based on child gender. The maternalquestioning, emotion knowledge association was strong and positive for girls. Forboys, the association was weak and negative. In other words, mothers’ greater use ofquestions during a wordless, emotionally salient storybook task predicts moreemotion knowledge for girls, but not for boys.

Secondary analyses

To account for the fact that missing data may be influencing the results, a parallel set ofanalyses were run in a structural equation modelling (SEM) framework using Amos

Table 2. Hierarchical regression analyses for predicting child emotion knowledge fromproportion of maternal questioning and child gender, controlling for age.

Variable B SE b R2 Radjusted2 F

Step 1 0.15 0.11 4.49∗

Age 0.43 0.21 0.38∗

Step 2 0.25 0.15 2.55∗∗

Age 0.47 0.21 0.42∗

Proportion of Questioning 9.96 18.34 0.11

Male 23.39 2.74 20.25

Step 3 0.35 0.24 3.11∗

Age 0.48 0.20 0.43∗

Proportion of Questioning 21.96 18.37 0.24

Male 22.72 2.61 20.20

Proportion∗Male 270.29 35.64 20.35∗∗

∗p , 0.05.∗∗p , 0.10.

Figure 1. Interaction between gender and low, average, and high proportion of questions onthe AKT.

Early Child Development and Care 273

Dow

nloa

ded

by [

Geo

rge

Mas

on U

nive

rsity

] at

12:

23 0

6 Ju

ne 2

013

(Arbuckle, 2003). These analyses use full information maximum likelihood (FIML)estimation to account for missing data. FIML has been shown to lead to less biasedestimates than list-wise deletion. The model was specified such that these analyseswere completely analogous to the regression analyses reported previously. Specifically,Age, Gender, Proportion of Questions, and the Gender X Proportion of QuestionsInteraction were entered as predictors of scores on the AKT. The model fit the datawell (x2 ¼ 3.02, df ¼ 3, p ¼ 0.39; comparative fit index ¼ 1.00, normed fit index¼ 0.95, root mean square error of approximation ¼ 0.009). In examining the parameterestimates, Age, b ¼ 0.41, Proportion of Questions, b ¼ 0.69, and the interactionbetween Gender and Proportion of Questions, b ¼ 20.55, were significant (p , 0.05)predictors of scores on the AKT, but Gender, b ¼ 20.03, alone was not. Theseresults mirror the results of the regression analyses (except that proportion was significantin this SEM model). Therefore, for the sake of parsimony, only results from theregression analyses will be discussed.

Discussion

Overall, we found that maternal questioning, after controlling for age, does not predict achild’s knowledge of emotions for all children. However, the association betweenmaternal questioning and emotion knowledge was found to be moderated by gender,such that daughters whose mothers asked more questions were more knowledgeableof emotions than daughters whose mothers asked fewer questions. By contrast, sonswhose mothers asked more questions were not necessarily more knowledgeable ofemotions than sons whose mothers asked fewer questions. These data support thenotion that maternal use of questions is an important component in how motherssocialise their daughters’ emotions.

Past research has given much support to argue the importance of parental socialisa-tion of emotion. For instance, parents give cues in their dialogue to help their childrendecipher emotional content (Baldwin & Meyer, 2007). Parents also explicitly talk aboutemotions (Denham & Auerbach, 1995; Denham et al., 1992), use frequent emotion talkwhen reacting to their child’s behaviour (Eisenberg & Fabes, 1994), and use emotiontalk in their own expressions and reactions to emotions to model emotional expression(Denham, Mitchell-Copeland, Strandberg, Auerbach, & Blair, 1997). Children are thuslearning to label their feelings, talk about and understand emotions, and appropriatelyexpress their emotions from their parents through emotion coaching (Gottman, Katz, &Hooven, 1996). Most importantly, not all parents ‘coach’ their children’s developmentof emotional competence equally, with differential child outcomes being explained bythis variability (Denham et al., 1997). What the present research revealed was that thevariability in mothers’ use of all questions, not just those about emotions, when readinga storybook filled with emotions predicted their daughters’ later knowledge ofemotions. Children’s days are filled with emotionally salient events and contexts.Parents’ use of questions during these events and in these contexts is a fertile timefor children to learn about emotions. Thus, the use of questions should be consideredwhen understanding parental socialisation of emotion.

Gender differences – mothers passing emotion knowledge to daughters

Scaffolding emotion knowledge using questions

Scaffolding is a concept that is linked with the Vygotskian theory stipulating that childrenare able to accomplish tasks when working with others when they would not be able to

274 C.S. Bailey et al.

Dow

nloa

ded

by [

Geo

rge

Mas

on U

nive

rsity

] at

12:

23 0

6 Ju

ne 2

013

accomplish it alone (Wood et al., 1976). There have been many conceptualisations ofwhat constitutes scaffolding (Benson, 1997). Benson (1997) notes that effective scaffold-ing often involves determining the level of the learner’s individual performance andasking leading questions that advance the learner from their individual performancetowards a higher level of understanding. Parents have been shown to use language,such as questions, to link previous knowledge with concurrent experiences (Morelock,Brown, & Morrissey, 2003). Questions within mother–daughter dyadic interactionserve to scaffold girls’ emotion knowledge in two ways. First, questions orient themother to the daughter’s thoughts and allow the mother to modify her behaviour basedon information that the daughter provides. Second, questions activate cognitive processesin the daughter that the mother feels are salient for the daughter’s understanding.

Questions have inherent ambiguity that prompts the listener to attend to the conceptbeing questioned about. Thus, when mothers ask their daughters, ‘Why do you thinkthe girl is sad?’, the daughter’s attention is directed to cues found within the story byshifting attention to the drawing of the girl and the possible antecedents that mightexplain the depiction of the character’s emotion expression. Baldwin and Meyer(2007) explain that language contains inherent emotional undertones and as such, thequestion in the preceding example contains cues for children (i.e. girl, sad), whichframe the eventual thinking about the emotional content. Furthermore, the cues inthe question help children unlock the other tidbits of information in the story (e.g.the puppy peed on the bed), eventually leading the child to answer the question.Social and emotional cues thus serve as a framework for interpreting the events forwhich the question was asked, and questions serve as a link between mothers anddaughters to interpret those cues. The ambiguity of the question prompts daughtersto think through the cues and antecedents to answer the question, strengthening theirunderstanding of causation and consequences of emotions. Parents can further usequestions to redirect when children’s answers are ‘incorrect’ or when the answer isunsatisfactory. Without questions, children may not know to which cues to attendand how to use cues to discover the antecedents and consequences of emotions. Inother words, mothers scaffold emotionally important information through questionsto aid their daughters in their future processing of social cues.

Why girls?

Our findings, which suggest that mothers ask more questions of girls than boys, is notentirely surprising given that mothers have been found previously to ask more ques-tions of girls (Cherry & Lewis, 1976; Clearfield & Nelson, 2006) and that mothersuse language to socialise emotions differently for boys and girls (Fivush et al., 2003;Kuebli et al., 1995). Although the present research found that mothers ask more ques-tions of girls than boys during a storybook laced with emotions, Crowley, Callanan,Tenenbaum, and Allen (2001) found that mothers ask more questions of boys thangirls during conversations about science. It is plausible that the source of the differen-tiated results may be related to gender stereotyping (Kennedy Root & Denham, 2010).In other words, mothers ask more questions of their daughters than sons when conver-sations or tasks involve emotions, reading, or interpersonal relationships, whereasmothers ask more questions of boys during conversations or tasks involving contentlike science. Altogether, maternal questioning during the emotionally laden storybookscaffolds young daughters’ processing of emotionally salient information for latersocial situations.

Early Child Development and Care 275

Dow

nloa

ded

by [

Geo

rge

Mas

on U

nive

rsity

] at

12:

23 0

6 Ju

ne 2

013

Our results may be explained by findings showing that mothers discuss more aboutemotions because of the social component of emotions (Fivush et al., 2003). Emotionsare embedded in the relationships that people and storybook characters have with othersand mothers may be ‘teaching’ their daughters about relating with others via knowingmore about emotions. Additionally, mothers may interact with their daughters differ-ently than their sons. Indeed, mothers and fathers do use more emotion words anddiscuss more about emotions with daughters than sons (Adams, Kuebli, Boyle, &Fivush, 1995; Fivush et al., 2003; Kuebli et al., 1995) and also model emotions differ-ently for daughters and sons (Garner et al., 1997). This means that boys and girls maybe learning about emotions differently. It is not that girls know more about emotionsthan boys (see Denham et al., submitted for publication), but that girls learn aboutemotions differently because of socialising processes in their environment. Mostlikely is that gender-specific socialisation processes are partly dependent on content,such as emotions, science, or interpersonal relationships, and partially dependent onthe method of socialisation, such as direct versus indirect.

Limitations

There are important caveats to our results. First, the design was not experimental and thedata are therefore correlational. Attempts to assign causality between questions andemotion knowledge cannot be made. Parents who ask more questions during emotionallyladen situations may cause children to be more knowledgeable about emotions but thepresent study cannot say so definitively. Second, after using list-wise deletion, asample size of 28 is generally considered low. However, the interaction, after controllingfor age and both main effects, is considered moderate. This demonstrates that samplesize need not be increased in order to detect the effect. Furthermore, the results werelargely replicated using FIML in an SEM framework, which lends support to thenotion that the small sample size was not masking a particular effect.

Future directions

The results from the current study reveal several potentially fruitful avenues of researchand practice worth exploring. First, future research can explore whether type of ques-tion leads to differential results. All questions were treated in the present study withequal weight in the analysis. No distinction was made based on the type of questionsasked. Some literature suggests that the value of asking questions is somewhat depen-dent on the features of the questions. For example, open-ended questions have differentcharacteristics and associated effects than close-ended questions (Blewitt et al., 2009).

Second, these analyses did not take child response into account. Developmentalphenomena are often theorised as transactional and the effect reported here is uni-directional, involving only half the dyad. The possibility that the children were elicitingquestions from their mothers may explain differences in gender described above. Inother words, mothers may ask proportionally more questions of children who havehigher rates of responding. This possibility is plausible given that girls initiated moreemotion-related discussions than boys and talked more about emotions than boys(Cervantes & Callanan, 1998; Kuebli et al., 1995).

Third, asking questions to scaffold social–emotional information can also beexplored in other populations, such as those of lower or higher socioeconomicstatus, who may have limited or additional access to books or similar opportunities

276 C.S. Bailey et al.

Dow

nloa

ded

by [

Geo

rge

Mas

on U

nive

rsity

] at

12:

23 0

6 Ju

ne 2

013

for dialogue about emotional situations. Also, it may be that the phenomenon operatesdifferently in special populations such as children with low understanding of emotionsor with fathers differently than with mothers.

Finally, children learn about emotions from the social world and policies need toreflect the growing literature that places environmental context at the forefront as towhy some children are better equipped for success in school because of their emotionalcompetence (Denham et al., in press). Furthermore, teachers may contribute to thesocialisation of emotion in children and their role in the development of emotionalcompetence cannot be ignored. As such, teachers can use conversation as a gatewayto discuss emotions and emotion situations in order to enhance children’s acquisitionof emotion knowledge. Teachers can also recognise the impact of their own expressionand reaction to emotions on children’s developing emotional competence.

Summary

The results reported here expand the work of Denham and Auerbach (1995) by dissect-ing a specific type of parent–child interaction. Children learn about emotions in thecontext of their relationship with others. The nature of this relation provides answersto questions of individual differences found in children’s knowledge of emotions.Asking questions when reading storybooks that involve many emotional situationsserves to scaffold social information about emotions, potentially paralleling the dialo-gues that parents have with their children. In particular, this study shows that girls maybe obtaining more value than boys, in terms of developing emotion knowledge, fromquestions asked by mothers about emotionally important information. Boys may belearning about emotions through different channels than girls.

Notes on contributorsCraig S. Bailey is an advanced graduate student at George Mason University. His researchfocuses on the socialisation of emotion and children’s development of emotional competencein classroom contexts.

Susanne A. Denham is an applied developmental psychologist and professor of psychology atGeorge Mason University. Her research focuses on the role of emotional competence in chil-dren’s social and academic functioning.

Timothy W. Curby is an assistant professor of psychology at George Mason University. Hisresearch focuses on teacher–student interactions and applying advanced statistical models toschool-based research.

ReferencesAdams, N.H. (1994, March). Ask, don’t tell: The value of asking young children questions.

Paper presented at the Association for Childhood Education International StudyConference, New Orleans, LA.

Adams, S., Kuebli, J., Boyle, P.A., & Fivush, R. (1995). Gender differences in parent-childconversations about past emotions: A longitudinal investigation. Sex Roles, 33, 309–323.

Arbuckle, J.L. (2003). Amos (Version 5) [Computer software]. Chicago, IL: SmallWaters.Baldwin, D., & Meyer, M. (2007). How inherently social is language? In E. Hoff & M. Shatz

(Eds.), Blackwell handbook of language development (pp. 87–106). Malden, MA:Blackwell.

Barr, R., Zack, E., Garcia, A., & Muentener, P. (2008). Infants’ attention and responsiveness totelevision increases with prior exposure and parental interaction. Infancy, 13, 30–56.

Early Child Development and Care 277

Dow

nloa

ded

by [

Geo

rge

Mas

on U

nive

rsity

] at

12:

23 0

6 Ju

ne 2

013

Benson, B.K. (1997). Scaffolding (coming to terms). English Journal, 86, 126–27.Blewitt, P., Rump, K.M., Shealy, S.E., & Cook, S.A. (2009). Shared book reading: When and

how questions affect young children’s word learning. Journal of Educational Psychology,101, 294–304.

Carr, D. (1998). The art of asking questions in the teaching of science. School Science Review,79, 47–60.

Cervantes, C.A., & Callanan, M.A. (1998). Labels and explanations in mother–child emotiontalk: Age and gender differentiation. Developmental Psychology, 34, 88–98.

Cherry, L., & Lewis, M. (1976). Mothers and two-year-olds: A study of sex-differentiatedaspects of verbal interaction. Developmental Psychology, 12, 278–282.

Chouinard, M.M. (2007). Children’s questions: A mechanism for cognitive development.Monographs of the Society for Research in Child Development, 72(1), 1–129.

Clearfield, M.W., & Nelson, N.M. (2006). Sex differences in mothers’ speech and play behaviorwith 6-, 9-, and 14-month-old infants. Sex Roles, 54, 127–137.

Cohen, J. (1992). A power primer. Psychological Bulletin, 112(1), 155–159.Cohen, P., Cohen, J., West, S.G., & Aiken, L.S. (2002). Applied multiple regression/correlation

analysis for the behavioral sciences (3rd ed.). New York, NY: Routledge Academic.Crowley, K., Callanan, M.A., Tenenbaum, H.R., & Allen, E. (2001). Parents explain more often

to boys than to girls during shared scientific thinking. Psychological Science, 12, 258–261.Denham, S.A. (2007). Dealing with feelings: How children negotiate the worlds of emotions and

social relationships. Cognitie Creier Comportament, 11, 1–48.Denham, S.A., & Auerbach, S. (1995). Mother-child dialogue about emotions and preschoolers’

emotional competence. Genetic, Social, and General Psychology Monographs, 121,313–337.

Denham, S.A., Bassett, H.H., Way, E., Kalb, S.C., Warren-Khot, H.K., & Zinsser, K. ‘Howwould you feel? What would you do?’ Development and underpinnings of preschoolers’social information processing. Manuscript submitted for publication.

Denham, S.A., Cook, M., & Zoller, D. (1992). ‘Baby looks very sad’: Implications of conversa-tions about feelings between mother and preschooler. British Journal of DevelopmentalPsychology, 10, 301–315.

Denham, S.A., & Couchoud, E.A. (1991). Social-emotional predictors of preschoolers’responses to adult negative emotion. Journal of Child Psychology and Psychiatry, 32,595–608.

Denham, S.A., McKinley, M., Couchoud, E.A., & Holt, R. (1990). Emotional and behavioralpredictors of preschool peer ratings. Child Development, 61, 1145–1152.

Denham, S.A., Mitchell-Copeland, J., Strandberg, K., Auerbach, S., & Blair, K. (1997). Parentalcontributions to preschoolers’ emotional competence: Direct and indirect effects. Motivationand Emotion, 21, 65–86.

Denham, S.A., Zinsser, K.M., & Brown, C.A. (In press). The emotional basis of learning anddevelopment in early childhood education. In B. Spodek & O. Saracho (Eds.), Handbookof research on the education of young children (3rd ed.). New York, NY: LawrenceErlbaum.

Eisenberg, N., Cumberland, A., & Spinrad, T.L. (1998). Parental socialization of emotion.Psychological Inquiry, 9, 241–273.

Eisenberg, N., & Fabes, R.A. (1994). Mothers’ reactions to children’s negative emotions:Relations to children’s temperament and anger behavior. Merrill-Palmer Quarterly, 40,138–156.

Fivush, R., Berlin, L., Sales, J.M., Mennuti-Washburn, J., & Cassidy, J. (2003). Functions ofparent-child reminiscing about emotionally negative events. Memory, 11, 179.

Garner, P.W., Carlson Jones, D., Gaddy, G., & Rennie, K.M. (1997). Low-income mothers’conversations about emotions and their children’s emotional competence. SocialDevelopment, 6, 37–52.

Goodman, L., & Berntson, G. (2000). The art of asking questions: Using directed inquiry in theclassroom. The American Biology Teacher, 62(7), 473–476.

Gottman, J.M., Katz, L.F., & Hooven, C. (1996). Parental meta-emotion philosophy and theemotional life of families: Theoretical models and preliminary data. Journal of FamilyPsychology, 10, 243–268.

278 C.S. Bailey et al.

Dow

nloa

ded

by [

Geo

rge

Mas

on U

nive

rsity

] at

12:

23 0

6 Ju

ne 2

013

Halberstadt, A.G., Denham, S.A., & Dunsmore, J.C. (2001). Affective social competence.Social Development, 10, 79–119.

Hart, C.H., Newell, L.D., & Olsen, S.F. (2003). Parenting skills and social-communicativecompetence in childhood. In J.O. Greene & B.R. Burleson (Eds.), Handbook of communi-cation and social interaction skills (pp. 753–797). Mahwah, NJ: Lawrence ErlbaumAssociates.

Honig, A.S., & Wittmer, D.S. (1981, August). Teacher questions to male and female toddlers.Paper presented at the Symposium of the International Society for the Study of BehavioralDevelopment, Toronto, Canada.

Kennedy Root, A., & Denham, S.A. (2010). The role of gender in the socialization of emotion:Key concepts and critical issues. In A.K. Root & S.A. Denham (Eds.), The role ofparent and child gender in the socialization of emotional competence (pp. 1–10) (NewDirections for Child and Adolescent Development). San Francisco, CA: Jossey-Bass.

Kuebli, J., Butler, S., & Fivush, R. (1995). Mother-child talk about past emotions: Relations ofmaternal language and child gender over time. Cognition & Emotion, 9, 265.

Kruper, J.C., & Uzgiris, I.C. (1987). Fathers’ and mothers’ speech to young infants. Journal ofPsycholinguistic Research, 16, 597–614.

Leaper, C. (2002). Parenting boys and girls. In M.H. Bornstein (Ed.), Handbook of parenting:Vol. 1. Children and Parenting. Mahwah, NJ: Erlbaum.

Leaper, C., Anderson, K.J., & Sanders, P. (1998). Moderators of gender effects on parents’ talkto their children: A meta-analysis. Developmental Psychology, 34, 3–27.

Lytton, H., & Romney, D.M. (1991). Parents’ differential socialization of boys and girls: Ameta-analysis. Psychological Bulletin, 109, 267–296.

Morelock, M.J., Brown, P.M., & Morrissey, A.-M. (2003). Pretend play and maternal scaffolding:Comparisons of toddlers with advanced development, typical development, and hearingimpairment. Roeper Review, 26, 41–51.

Saarni, C., Mumme, D.L., & Campos, J.J. (1998). Emotional development: Action, communi-cation, and understanding. In N. Eisenberg (Ed.), Handbook of child psychology: Social,emotional, and personality development (5th ed., Vol. 3, pp. 237–309). Hoboken, NJ:John Wiley & Sons.

Tenenbaum, H.R., & Leaper, C. (2002). Are parents’ gender schemas related to their children’sgender-related cognitions? A meta-analysis. Developmental Psychology, 38, 615–630.

Vygotsky, L.S. (1978). Mind in society: The development of higher psychological processes(M. Cole, V. John-Steiner, S. Scribner, & E. Souberman, Eds.). Cambridge, MA: HarvardUniversity Press.

Warren, H.K., & Stifter, C.A. (2008). Maternal emotion-related socialization and preschoolers’developing emotion self-awareness. Social Development, 17, 239–258.

Weinstein, Y., McDermott, K.B., & Roediger, H.L. (2010). A comparison of study strategiesfor passages: Rereading, answering questions, and generating questions. Journal ofExperimental Psychology: Applied, 16, 308–316.

Winsler, A., Fernyhough, C., McClaren, E.M., & Way, E. (2005). Private speech codingmanual. Unpublished manual. George Mason University, Fairfax, VA. Retrieved fromhttp://classweb.gmu.edu/awinsler/Resources/PsCodingManual.pdf

Wood, D., Bruner, J.S., & Ross, G. (1976). The role of tutoring in problem solving. Journal ofChild Psychology and Psychiatry, 17, 89–100.

Early Child Development and Care 279

Dow

nloa

ded

by [

Geo

rge

Mas

on U

nive

rsity

] at

12:

23 0

6 Ju

ne 2

013