testing the vision: preschool settings as places for meeting, bonding and bridging

13
Testing the Vision: Preschool Settings as Places for Meeting, Bonding and Bridging Karen Thorpe*, Sally Staton and Robert Morgan School of Psychology and Counseling, Queensland University of Technology, Brisbane, Qld, Australia Susan Danby School of Early Childhood, Queensland University of Technology, Brisbane, Qld, Australia Collette Tayler Graduate School of Education, University of Melbourne, Melbourne, Vic., Australia The OECD (2006 Starting Strong II: Early Childhood Education and Care. OECD Publishing: Paris) envisions early childhood education and care settings as meeting places for diverse social groups; places that build social capital. This vision was assessed in a comparison of three preschools types: full-fee paying, subsidised-fee and publicly funded. The social compo- sition within each was examined and the connectedness of the children (n = 472) who attended compared. Publicly funded preschools had more socially diverse populations. The quantity of social connectedness did not differ but children in publicly funded preschools described higher quality social relationships. Not all preschool settings are socially diverse but, where they are, the quality of relationships is highest. Ó 2010 The Author(s). Children & Society Ó 2010 National Children’s Bureau and Blackwell Publishing Limited. Keywords: preschool, relationship quality, social capital, social connectedness, social inclusion, sociometrics. Converging evidence from the economic and developmental sciences has placed focus on the early years as foundational in establishing pathways to personal productivity and social par- ticipation. Considerable attention has been focused on human capital formation and, particu- larly, the long-term educational benefits deriving from high quality early education and care (ECEC) (Reynolds and Temple, 2008; Schweinhart and others, 2005). Alongside, there has been interest in social capital formation: the potential for early education settings to forge social connectedness and shared values. The literature identifies two key forms of social cap- ital (Almedom, 2005; Baum and Ziersch, 2003). The first, bonding capital, is the degree of connectedness and trust within a social group and serves to provide protection and security. The second, linking capital, refers to connectedness and mutual trust across social groups and is associated with individual social advancement and broader social harmony. It is this latter form that has been of particular theoretical interest and a focus for policy (Coleman, 1990; Council of Australian Governments, 2009). In Starting Strong II, the OECD (2006) identifies the potential for ECEC settings to connect individuals across the diversity of society. Throughout, the report asserts the right of all chil- dren to participate in quality ECEC programmes and argues that ECEC delivers a public good that includes ‘social cohesion’ (p. 37). The concluding chapter (p. 219) presents an aspiration CHILDREN & SOCIETY VOLUME 26, (2012) pp. 328–340 DOI:10.1111/j.1099-0860.2010.00336.x Ó 2010 The Author(s) Children & Society Ó 2010 National Children’s Bureau and Blackwell Publishing Limited

Upload: karen-thorpe

Post on 25-Sep-2016

215 views

Category:

Documents


1 download

TRANSCRIPT

Page 1: Testing the Vision: Preschool Settings as Places for Meeting, Bonding and Bridging

Testing the Vision: Preschool Settings asPlaces for Meeting, Bonding and BridgingKaren Thorpe*, Sally Staton and Robert MorganSchool of Psychology and Counseling, Queensland University of Technology, Brisbane, Qld, Australia

Susan DanbySchool of Early Childhood, Queensland University of Technology, Brisbane, Qld, Australia

Collette TaylerGraduate School of Education, University of Melbourne, Melbourne, Vic., Australia

The OECD (2006 Starting Strong II: Early Childhood Education and Care. OECD Publishing:

Paris) envisions early childhood education and care settings as meeting places for diverse

social groups; places that build social capital. This vision was assessed in a comparison of

three preschools types: full-fee paying, subsidised-fee and publicly funded. The social compo-

sition within each was examined and the connectedness of the children (n = 472) who

attended compared. Publicly funded preschools had more socially diverse populations. The

quantity of social connectedness did not differ but children in publicly funded preschools

described higher quality social relationships. Not all preschool settings are socially diverse

but, where they are, the quality of relationships is highest. � 2010 The Author(s). Children &

Society � 2010 National Children’s Bureau and Blackwell Publishing Limited.

Keywords: preschool, relationship quality, social capital, social connectedness, socialinclusion, sociometrics.

Converging evidence from the economic and developmental sciences has placed focus on theearly years as foundational in establishing pathways to personal productivity and social par-ticipation. Considerable attention has been focused on human capital formation and, particu-larly, the long-term educational benefits deriving from high quality early education and care(ECEC) (Reynolds and Temple, 2008; Schweinhart and others, 2005). Alongside, there hasbeen interest in social capital formation: the potential for early education settings to forgesocial connectedness and shared values. The literature identifies two key forms of social cap-ital (Almedom, 2005; Baum and Ziersch, 2003). The first, bonding capital, is the degree ofconnectedness and trust within a social group and serves to provide protection and security.The second, linking capital, refers to connectedness and mutual trust across social groupsand is associated with individual social advancement and broader social harmony. It is thislatter form that has been of particular theoretical interest and a focus for policy (Coleman,1990; Council of Australian Governments, 2009).

In Starting Strong II, the OECD (2006) identifies the potential for ECEC settings to connectindividuals across the diversity of society. Throughout, the report asserts the right of all chil-dren to participate in quality ECEC programmes and argues that ECEC delivers a public goodthat includes ‘social cohesion’ (p. 37). The concluding chapter (p. 219) presents an aspiration

CHILDREN & SOCIETY VOLUME 26, (2012) pp. 328–340DOI:10.1111/j.1099-0860.2010.00336.x

� 2010 The Author(s)

Children & Society � 2010 National Children’s Bureau and Blackwell Publishing Limited

Page 2: Testing the Vision: Preschool Settings as Places for Meeting, Bonding and Bridging

for early childhood centres that support social inclusion, social participation, social under-standing and democratic process. Social capital formation is presented as one of the keyvisions:

... the early childhood centre becomes a space where the intrinsic value of each person is recognised,where democratic participation is promoted, as well as respect for our shared environment. Learningto be, learning to do, learning to learn and learning to live together should be considered as criticalelements of each child toward human and social development. (Executive summary, p. 18)

Underlying this vision are two implicit hypotheses about the functioning of ECEC. The firstis that ECEC settings are accessed by the diverse range of social groups and thereby providesocial meeting places. The second is that early educational settings facilitate social connect-edness and mutual values. The current study empirically assesses these two hypotheses bycomparing three types of preschool provision. These are comparable in both business model(non-profit) and structural provision (class size, programme hours, teacher qualification), buthave different funding structures (full fee paying, subsidised reduced fee and no-fee). Thestudy design examines the social composition within these different preschool provisions andcompares the social connectedness of the children who attend.

Hypothesis 1: Early educational settings are social meeting places

Early childhood contexts, and therefore social groups in which young children are placed, aredetermined by family choices and are not random (Ball and Vincent, 1998; Prieto and others,2002; Thorpe and others, 2004; Vincent and Ball, 2006). Rather, these choices are associatedwith a number of factors: availability of programmes within the selected residential commu-nity (Prieto and others, 2002; Thorpe and others, 2004), accessibility and affordability of pro-grammes (Prieto and others, 2002; Thorpe and others, 2004), patterns of parent participationin paid employment (Thorpe and others, 2004) and family and community attitudes to differ-ent programmes (Ball and Vincent, 1998; Prieto and others, 2002; Vincent and Ball, 2006). Aconsequence is that the social composition of an ECEC setting may not always be diverse.

There is evidence that the mechanisms of ‘free market’ in parent choice of programme canresult in social homogeneity. A Swedish study presents evidence that parent choices of earlychildhood programme result in economic, social and cultural segregation (Prieto and others,2002). This detailed study of the social geography of childcare usage in the Swedish city ofUppsala found that the social pool of children in childcare settings did not relate to theirgeographical location of residence alone. While place of residence was itself socially segre-gating, differences in patterns of parent choice between socially advantaged and disadvan-taged groups further increased segregation. High income families more often chose privatepreschool settings that were not accessible for lower income families. The findings of thisstudy indicate that higher income families have greater choice, and that the basis of thischoice is different from that of lower income families.

An Australian study of early education provides further evidence of this process. Thorpe andothers (2004) questioned 1860 parents across 39 diverse communities about the reason fortheir choice of education setting for their four- to five-year-old children. The study foundthat the quality and reputation of the programme was important for all parents. However,for lower income families, pragmatic considerations (programme covering parent work hours,proximity to home ⁄ work, siblings on same site, cost) were significantly more salient in their

ECEC and Social Capital 329

� 2010 The Author(s) CHILDREN & SOCIETY Vol. 26, 328–340 (2012)

Children & Society � 2010 National Children’s Bureau and Blackwell Publishing Limited

Page 3: Testing the Vision: Preschool Settings as Places for Meeting, Bonding and Bridging

choice than they were for higher income families. Higher income families were more able totravel to obtain the programme they wanted and to pay for additional carers who couldfacilitate access to part-time preschool programmes during their work hours. High incomefamilies also appeared more able to advocate for their children to obtain enrolment in themost highly sought programmes.

Differences in parent choice reflect attitudinal differences and community values as well asdifferential availability, access and advocacy. In a Swedish study (Prieto and others, 2002),higher income families living in lower income locations were less likely to choose privateschools than high income families living in high income areas. A British sociological studyof parental choice of childcare and early education supports this view (Ball and Vincent,1998; Vincent and Ball, 2006). Ball and Vincent (1998) report that the reputation of differenteducation and care provision is directed through the ‘grapevine’ of personal communicationwithin a neighbourhood community. Thus while parental choice can result in social homoge-neity within ECEC settings this need not be the case.

In the current study of 24 preschool centres located in Brisbane, Australia, we focus on thefunding structure of the preschool settings as a factor affecting parental choice and socialaccess. We examine the social composition of preschools within the three common typesavailable: full-fee paying, subsidised-fee and publicly funded. All three types were subject tothe same regulation regime under the auspices of the Department of Education and, there-fore, comparable on potentially confounding structural features (staff ratios, teacher qualifi-cations and age of cohort) but varied on economic accessibility. In the first, full fee-payingprivate preschools, parents pay the full cost of provision. In the second, the cost is sharedbetween parents and the state such that the state covers the cost of teacher’s salaries. In thethird, there are no direct costs to parents, with the provision fully funded by the state. Thestudy specifically excluded childcare settings because, at the time of this study,1 such provi-sion was regulated by a separate authority and subject to different business and structuralregulations. These were potentially confounding and rendered childcare centres inappropriatefor comparison.

Hypothesis 2: Early educational settings facilitate social connectedness

The literature clearly indicates that parental choice works to influence the level of socialheterogeneity in ECEC settings and, thereby, the opportunity for social bridging. What is cur-rently unclear is how the level of heterogeneity might impact on children’s social connected-ness to each other and the underlying basis for this connection. The evidence is that wherethere is greater parental choice (and this occurs where there is greatest economic and socialcapital), there the social composition of the group will be more homogeneous. Higher levelsof similarity would presuppose opportunity for bonding capital and, therefore, stronger rela-tionships in less socially diverse settings. In contrast, current social and educational policytheory, such as that proposed by the OECD (2006), focuses on convergence of differencewithin the early childhood setting. This predicts a stronger social connectedness in settingsthat are socially heterogeneous. The current study examines these competing hypotheses byexamining the quantity and quality of social connections between peers in preschools withvarying social access.

Early education centres most often provide the first extra-familial social experiences andopportunities for social learning through interactions with peers. While studies of children’s

330 Karen Thorpe et al.

� 2010 The Author(s) CHILDREN & SOCIETY Vol. 26, 328–340 (2012)

Children & Society � 2010 National Children’s Bureau and Blackwell Publishing Limited

Page 4: Testing the Vision: Preschool Settings as Places for Meeting, Bonding and Bridging

friendship patterns suggest that children, like adults, have a preference to interact with thoseof similar race, age and interest to themselves (Aboud and Mendelson, 1996; Haselager andothers, 1998; Thorpe and Gardner, 2006), there is also evidence that early educational expe-riences can work to forge relationships across diverse groups. A range of studies examiningthe social inclusion of children from different cultures (Howes and Wu, 1990; Thorpe andothers, 2009), with different levels of ability (Buysse and others, 2002) or with differentlevels of social advantage (Sylva and others, 2004), attest to the potential opportunity forsocial bridging.

One particular feature of ECEC settings identified by research as serving to bridge socialdivides is the opportunity for social interaction through play (e.g. Buysse and others,2002). The study of social inclusion of children with disability provides a good example. Aseries of studies examining the formation of friendship among children in integratedclasses, composed of both normally developing children and those with disability, indicatethat the opportunity to interact in play was associated with the number (Buysse and others,2002, 2003a) and quality (Guralnick and others, 1996) of friendships formed in bothgroups.

However, the provision of time for play does not necessarily facilitate social connection.This is particularly the case where children have impediments to their social access. Oneform of such impediment relates to child characteristics. Thus studies of twin children,who more commonly have poorer language development (Thorpe, 2006), have found thatprovision of opportunities for play in a preschool setting does not necessarily result inimproved social skills or engagement with peers (Hay and Preedy, 2006; Thorpe, 2003a).Similarly, children with disability may have reduced access to social interaction within anintegrated classroom (Buysse and others, 2002, 2003b). A second form of impediment tosocial access relates to class group characteristics. A year-long observational study of atargeted programme for disadvantaged children found that the quality of play and amountof verbal interaction actually diminished over the year (Farran and Son-Yarbrough, 2001).Though the structural features and observed quality of the programme was high in thisstudy, a key feature was the social homogeneity of the children who attended. Entry to theprogramme was based on screening for educational difficulty within a highly disadvan-taged population. The findings of this study align with reports from the Effective Provisionof Early Education Study (Sylva and others, 2004) that indicate that socially disadvantagedchildren performed better when they attend socially diverse preschools. Together, thesestudies suggest that class group composition affects access to social learning opportunitiesand social connectedness of children.

The current study asked whether connectedness and quality of relationship is higher in set-tings that are more or less socially heterogeneous. It assessed the quantity and quality ofconnectedness across different preschool settings. Social connectedness was assessed throughsociometric measurement while quality of relationships was assessed through interviews withthe children in which they gave accounts of their relationships with the peers nominated inthe sociometric task. This approach takes the theoretical perspective that children are compe-tent informants about their own experiences (Danby and Farrell, 2004; James and Prout,1997) and is a response to the growing calls for children’s accounts of social capital (Daviesand others, 2007; Farrell and others, 2004; Morrow, 2001). Currently there are few examples(Farrell and others, 2004; Morrow, 2001).

ECEC and Social Capital 331

� 2010 The Author(s) CHILDREN & SOCIETY Vol. 26, 328–340 (2012)

Children & Society � 2010 National Children’s Bureau and Blackwell Publishing Limited

Page 5: Testing the Vision: Preschool Settings as Places for Meeting, Bonding and Bridging

Method

Sample

Class groups were those attended by a twin child participating in the study of twin children’ssocial competence. Twins were recruited through the Australian Multiple Birth Association,the Australian Twin Registry and advertisements. Single-born controls were selected fromthe class attendance register by taking the next child on the classroom enrolment list of thesame sex as the twin child. Data on children’s demographics, behaviour, language skills andfriendships were collected from the teacher and parent on the twin and control children. Inaddition, all classmates were invited to participate in the sociometric study. Sociometric datawere collected from all classmates who provided parental consent. The sample was drawnfrom a potential pool of 544 enrolled children, of which twin children comprised 9.9% of thesample. The children attended 24 different pre-school classes, six private preschool (n = 103;male = 51, female = 52), four publicly subsidised preschool (n = 72; male = 36, female = 36)and 14 fully publicly funded preschool (n = 297; male = 152, female = 145). Sociometricdata were completed by 472 preschoolers yielding a participation rate of 87 per cent. Ageand demographic data were only available for the subset of children (twins and controls)who were participating in the transition to school study (n = 96; 50 twin children; 46 single-tons). The mean age of this sub-group was 60.39 months (males: mean = 60.89, SD = 4.38;females: mean = 59.96, SD = 4.14).

Measures and procedures

Social networkSocial network was measured using a modification of the ‘The Bus Story’ procedure reportedby Perren and Alsaker (2006). In the original procedure children were asked to select class-mates from their kindergarten group to go on a bus adventure. Children made their selectionby choosing from photographs of all children in the class. In our procedure, the concept ofgoing on an adventure was also used, but children worked with researchers to select theadventure and build the story. Drawings rather than photographs were used to conform toethical committee requirements that did not permit the use of photography. Children wereasked to nominate three children. The Bus Story was completed in the classroom setting andwas explained to children in small groups (2–4 children). Each child individually completedtheir drawing identifying their three nominated friends and was interviewed by a researcherabout the reasons for their nomination.

The following protocol was used:

To find out about your friends we are going to ask you to think about who you would take withyou on an adventure in this bus (show bus). Where would you choose to go if you could go on anadventure on this bus (child names an adventure ⁄ place they would like to go). You can choose threefriends to take with you. Are you happy to draw your friends? (Only consenting children continue)

In addition, children were asked about the reasons for their selection. Their responses weretranscribed verbatim. No limit was placed on the number of reasons that a child could pres-ent. The protocol continued as follows:

332 Karen Thorpe et al.

� 2010 The Author(s) CHILDREN & SOCIETY Vol. 26, 328–340 (2012)

Children & Society � 2010 National Children’s Bureau and Blackwell Publishing Limited

Page 6: Testing the Vision: Preschool Settings as Places for Meeting, Bonding and Bridging

Who is this friend?Why did you choose to take him ⁄ her?Tell me about (child’s name). Why is he ⁄ she your friend?

A coding framework was developed to categorise these data on the basis of prior literatureand examination of the empirical data. This categorised friendship quality into five levelsbased on evidence of the degree of longevity, emotional affiliation and reciprocation of achild’s friendship:

d Here & Now: reasons provided are immediate, short term or a simple restatement. Evidenceof duration is absent. For example, ‘We play’, ‘I like him ⁄ her’.

d Physical: reasons given are based on the physical attributes such as appearance, physicalskill or ownership of an object. For example, ‘She ⁄ he has nice hair’, ‘She ⁄ he is good atrunning’, ‘She ⁄ he has a nice pencil case’.

d Personal: reasons given are based on personal attributes and particularly those pertainingto emotion and relationship such as humour, kindness, sharing. For example, ‘She ⁄ healways makes me laugh’, ‘She ⁄ he always helps me’. Also included here were the responsesin which the absence of negative behaviour was given as reasoning. For example, ‘She ⁄ hedoes not hurt me’.

d History: reasons given are based on a shared history with other children over time andextending beyond school For example, ‘We have known each other a very long time’, ‘Ourparents are friends’, ‘We always go to each other’s house’.

d Best friend: reasons given are based on child nominating another as their best friend ⁄ s.The use of these terms sit alongside reasons suggesting reciprocation. For example,‘She ⁄ he’s my best friend’, ‘We always play together’.

Inter-rater reliability analyses using two independent raters on a 15 per cent sample (n = 79)were undertaken. Interclass correlations applying gamma correlations were: Here and NowC = 0.69, Physical C = 0.90, Personal C = 0.83, History C = 1.0, Best friend C = 0.96 (Siegeland Castellan, 1988).

From this procedure three key variables were derived: (1) Number of Reciprocal Nominationswhich is a measure of mutual friendship, (2) Number of Reasons given for each Nominationwhich is a measure of the multiple qualities of the relationship and (3) Highest Order Reasonused to Explain Nomination which is a measure of the endurance and emotional affiliationof the relationship.

Child language and behaviourLanguage and behaviour problems are potential impediments to the formation and mainten-ance of peer relationships. Though this study did not examine the effects of languageand behaviour on social relationships, these variables presented potential confoundingexplanations if these characteristics varied systematically across the three preschool types.The language measure was a modification of the Bates Macarthur Communicative Develop-ment Inventory undertaken by Dale and others (2000) to extend its use for four-year-olds.The measure included a measure of language complexity and pragmatics, and from this asingle language usage score was derived. The behaviour measure was the Strengths andDifficulties Questionnaire (Goodman, 1997). For this study the total difficulties score wasused.

ECEC and Social Capital 333

� 2010 The Author(s) CHILDREN & SOCIETY Vol. 26, 328–340 (2012)

Children & Society � 2010 National Children’s Bureau and Blackwell Publishing Limited

Page 7: Testing the Vision: Preschool Settings as Places for Meeting, Bonding and Bridging

Social indexTo measure social status an index was derived based on parent report of highest level ofeducation for father and mother and gross annual income for family. An indexing procedurewas used to scale the three measures into a single aggregate score. Highest level of educationfor each parent was assigned scores weighted from 0 to 2 (0 = up to year 10, 1 = year 12 ordiploma certificate, 2 = university education). For single parent families a mean computationwas entered. Gross annual income was likewise weighted based on normal income distribu-tion for the sample (0 = $0–$60 000, 1 = $61 000–$100 000, 2 = over $100 000). Aggre-gated weightings were used to form three index groups: high, medium and low social status.

Child-initiated playAs part of the study of school transitions and social competence, children from the partici-pating groups were video-recorded as they interacted in their early childhood centre. Thepurpose of the video-recording was to examine children’s competencies with peers in a playsetting. Teachers therefore were asked to identify the most appropriate time to access play(defined as child-initiated activity) in their classrooms and visits were scheduled accordingly.The visits were undertaken by two researchers who subsequently independently rated theclassroom observations (those on tape and those observed in the visit) according to thefrequency and duration of opportunity for child-directed activity. The rating was made on athree-point scale: 1 (some opportunity), 2 (frequent opportunity), 3 (extended opportunity).Ratings had 100 per cent agreement.

Results

Comparability of sample

Mean participation rates ranged from 81 per cent for private to 90 per cent for publicallyfunded, and were not significantly different across preschool types. To examine the possibilityof selection bias associated with age, language and behavioural characteristics of the childrenacross the three preschool types (Table 1), group comparison was made. ANOVA indicatedthat there were no statistical differences in mean age [F(2,93) = 0.416, P = 0.661], language[F(2,93) = 1.95, P = 0.148] or behavioural difficulty scores [F(2,93) = 1.04, P = 0.359].

Hypothesis 1: Social distribution of the sample

Figure 1 shows the distribution of family social index data, based on highest education levelof each parent and gross family income, across each of the three preschool types. Preschoolsthat were publically funded showed the most even distribution of family social index groupswith almost equal distribution representing each of low (34.3%), medium (31.4%) and high

Table 1: Mean (SD) sample characteristics by preschool type

Preschool type

TotalPrivate Publicly subsidised Publicly funded

Age (months) 59.6 (4.1) 61.2 (5.6) 60.7 (4.1) 60.4 (4.3)Language score 50.1 (9.9) 51.1 (8.16) 45.9 (11.0) 47.3 (10.6)Total difficulties score 15.7 (3.5) 18.0 (5.1) 15.8 (5.0) 16.1 (4.8)

334 Karen Thorpe et al.

� 2010 The Author(s) CHILDREN & SOCIETY Vol. 26, 328–340 (2012)

Children & Society � 2010 National Children’s Bureau and Blackwell Publishing Limited

Page 8: Testing the Vision: Preschool Settings as Places for Meeting, Bonding and Bridging

(34.3%) groups. Preschools that were publicly subsidised had an equally high proportion ofthose from high (42.9%) and medium (42.9%) groups, and less from the lower social indexgroup (14.3%). The distribution of family social index across private preschools showed askewed pattern towards a greater proportion of those from high (50%), than medium (30%)and low (20%) social groups.

Hypothesis 2: Social networks and qualities of friendships across preschools

The descriptive statistics for the friendship data are presented in Table 2. To account for non-homogeneous sample variances and ordinal data, group differences in the number and highestorder reasons for nominations were examined via nonparametric Kruskal–Wallis and Mann–Whitney analyses. Results indicated a significant differences across the three preschool typesin the number of reasons given by the children for their nominations, v2 (2,472) = 15.716,P < 0.001. Follow-up Mann–Whitney tests, applying a Bonferonni adjustment, indicated thatchildren in publicly funded preschools (mean rank = 195.3) gave significantly more reasonsfor their peer nominations than children in publicly subsidised settings (mean rank = 142.8),U = 7652.000, P < 0.001. There was no significant difference between private and publicallyfunded or publically subsidised preschool types. There was also a significant difference in thehighest order reason given by children for their nominations across the pre-school types,

0

10

20

30

40

50

60

Private preschool Publicly subsidizedpreschool

Publicly fundedpreschool

Per

cen

tag

e o

f fa

mili

es

Preschool type

Social index: Low

Med

High

Figure 1: Distribution of social status within private, publicly subsidised and publicly funded preschoolgroups.

Table 2: Descriptive statistics for friendship measures by preschool type

Preschool type

TotalPrivate Publicly subsidised Publicly funded

Number of reciprocal nominationsMean (SD) 0.78 (0.86) 0.90 (0.79) 0.86 (0.82) 0.83 (0.83)Range 0–3 0–3 0–3 0–3

Number of reasons given for nominationMean (SD) 4.05 (1.71) 3.74 (0.98) 4.48 (1.87) 4.27 (1.75)Range 0–9 3–6 0–11 0–11

Highest order reasonMean (SD) 2.16 (1.41) 2.46 (1.57) 2.63 (1.56) 2.48 (1.54)Range 0–5 0–5 0–5 0–5

ECEC and Social Capital 335

� 2010 The Author(s) CHILDREN & SOCIETY Vol. 26, 328–340 (2012)

Children & Society � 2010 National Children’s Bureau and Blackwell Publishing Limited

Page 9: Testing the Vision: Preschool Settings as Places for Meeting, Bonding and Bridging

v2 (2,472) = 7.579, P = 0.023. Post hoc analysis, using a Bonferonni adjustment, showed thatchildren in publicly funded preschools (mean rank = 209.5) gave reasons of significantlyhigher quality for their nominations than their private preschool counterparts (meanrank = 174.6), U = 12626.5, P < 0.006. There was no significant difference between publicallysubsidised and publically funded or private preschool groups. The number of reciprocal nomi-nations did not differ significantly across the three preschool types.

Preschool type, child-initiated play and friendshipMean access to child-initiated play scores are presented in Table 3. Because ratings of child-initiated play were ordinal data rather than continuous, nonparametric tests were undertaken.Kruskal–Wallis analysis revealed significant differences between the preschool groups in theamount of access to child-initiated free play [v2(2,472) = 232.123, P < 0.001]. Publicly sub-sidised settings were scored as providing significantly more access to child-initiated free playthan both private (U = 576.000, P < 0.001) and publicly funded (U = 9324.000, P = 0.001)preschool settings. There was also a significant difference between publicly funded and pri-vate preschool settings, with public settings scored as providing significantly more access tochild-initiated free play (U = 4149.500, P < 0.001).

The association of access to child initiated play and children’s relationship outcomes werealso investigated using Spearman rank correlations. Small but significant correlations werefound between access to child-initiated free play and the number of reciprocated friendships(rho = 0.144, P = 0.002), highest order reason (rho = 0.104, P = 0.024) and the number ofreasons given for nominations (rho = 0.118, P = 0.01).

Discussion

There has been considerable theoretical interest in ECEC as a vehicle for social capital forma-tion (Council of Australian Governments, 2009; OECD, 2006). However, to date there hasbeen little empirical testing of the role ECEC plays in bridging social divides. The currentstudy addressed two key research hypotheses in this respect:

d That early educational settings are social meeting placesd That early educational settings facilitate social connectedness

The approach for investigation involved obtaining both quantitative sociometric data andqualitative accounts about social connectedness with peers from preschool children. Bothapproaches are a response to the call for research directly involving children in reporting ofsocial capital and also address our concern that sociometric data alone provide insufficient

Table 3: Descriptive statistics for child-initiated play by preschool type

Mean SD Mean rank Range

Private 1.73 0.72 97.88 1–3Publicly subsidised 3.00 0.00 299.00 3Publicly funded 2.82 0.50 269.42 1–3Total 2.61 0.70 n ⁄ a 1–3

Range: 1 = some, 2 = frequent, 3 = extend.

336 Karen Thorpe et al.

� 2010 The Author(s) CHILDREN & SOCIETY Vol. 26, 328–340 (2012)

Children & Society � 2010 National Children’s Bureau and Blackwell Publishing Limited

Page 10: Testing the Vision: Preschool Settings as Places for Meeting, Bonding and Bridging

insight into the quality of social connections. The analyses compared quality of social con-nections, both mutual nominations and qualitative accounts, across three preschool typeswith different levels of social access. Additionally they assessed the association withobserved pedagogical practice within the different types.

Initial analyses assessed the comparability of the samples in each of the three preschooltypes to ensure that any finding was not an artefact of systematic variations in the individ-ual characteristics of children attending. Although data were not available on all children inthe sample, analyses of the randomly selected groups within each class suggested that thesamples within each preschool type were comparable in key characteristics that may affectsocial interaction: age, language skill and behaviour. Further, the participation rates did notvary across settings.

Hypothesis 1: Early educational settings are social meeting places

The levels of social heterogeneity across the three groups varied. There was a marked under-representation of children from socio-economically disadvantaged families in both publiclysubsidised and private (full-fee) paying preschools. Fully publicly funded preschools, in con-trast, were socially heterogeneous and had an even social distribution as indexed by the com-bined measure of parental education and income. These findings are consistent with priorstudies (Prieto and others, 2002; Thorpe and others, 2004; Vincent and Ball, 2006) and indicatethat while economic barriers produce social homogeneity in fee-paying settings, they do notcommensurately preclude social heterogeneity in those that are free of charge. Many sociallyadvantaged parents in our sample chose to place their children in a publicly funded preschool.

Hypothesis 2: Early educational settings facilitate social connectedness

Having established that different preschool types under investigation varied in their level ofsocial heterogeneity, we sought to examine whether this was systematically associated withthe children’s reports of social connectedness. Our results indicated that there were no differ-ences in the number of reciprocated friendships across the three preschool types. That is,regardless of the level of social heterogeneity of the setting, the children were equally ableto establish a reciprocated relationship. What is particularly notable about this finding is thatchildren in the social heterogeneous setting were no less likely to establish a reciprocatedrelationship. This is an important finding, as reciprocated friendships are reported to havegreater developmental significance than interactions with children who do not share mutualaffiliation (Berndt, 2004; Newcomb and Bagwell, 1995; Rubin and others, 2006).

The most evident differences to emerge from the data were those pertaining to the children’squalitative descriptions of their social relationships. The number of reasons given for peernomination and the quality of relationship signified by these descriptions varied systematicallyacross the preschool settings. The number of reasons given for a peer relationship was highestin publicly funded settings. This difference was statistically significant when compared withpublicly subsidised preschools but did not attain a level of statistical significance for privatepreschools. The quality of the peer relationship was also highest in publicly funded preschoolsbut, on this measure, lowest in the private preschools. That is, while children in privatepreschools presented multiple reasons for their friendships, and as many reasons as thosepresented by children in publicly funded settings, they were less likely to describe relationshipsthat were enduring, reciprocated and emotionally attached.

ECEC and Social Capital 337

� 2010 The Author(s) CHILDREN & SOCIETY Vol. 26, 328–340 (2012)

Children & Society � 2010 National Children’s Bureau and Blackwell Publishing Limited

Page 11: Testing the Vision: Preschool Settings as Places for Meeting, Bonding and Bridging

One explanation for these findings is that the opportunities to engage in social interactionsthrough play differed across the three settings (see Buysse and others, 2002, 2003a;Guralnick and others, 1996). Certainly the opportunity for child-initiated play was high inthe publicly funded preschools where the quality of social relationship was also greatest.However, opportunities for child-initiated play were equally high in the publicly subsidisedsettings where relationship quality was not correspondingly high. The opportunities for socialinteraction through play present at best only a partial explanation for these findings thoughwarrant further exploration with more refined measurement and more extensive observationthan that undertaken in the current study.

A second explanation lies in the social composition of the class group. Our data suggest thatchildren in more heterogeneous preschool settings experience higher quality relationshipsthan children in settings that are less socially diverse. These findings accord with priorresearch that report better outcomes for children in socially heterogeneous groups (Sylva andothers, 2004) and regression of social interaction for children placed in highly homogenous,disadvantaged groups (Farran and Son-Yarbrough, 2001). What is important about thecurrent study is that it extends these findings to report poorer social outcomes forchildren placed in homogenously advantaged groups.

Our study was able to report on the quality of peer relationships across three preschool typesdistinguished by the heterogeneity of their social composition. However, a limitation of thestudy was that it could not examine patterns of social bridging within each classroom. Becausedemographic and social index data were only available for a random subset of each class, wewere unable to explicitly assess social bridging by examining the social status of each child ina nomination dyad. In a prior study of a racially mixed classroom in which full demographicdata were available we were able to confirm cross-race friendships and track their develop-ment across time (Thorpe and others, 2009). Future studies designed expressly to examinehomogeneity effects should include detailed demographic and social data and track effectsacross time. Such approaches provide outcome indices for investigation and experimentationof policy and pedagogical practices that foster bridging across diverse social groups.

The study also directs attention to the explanation of mechanisms that underlie the positiveeffect of social heterogeneity. The work of developmental scientists that examines the devel-opment of social understanding and theory of mind provides a fruitful direction. Research inthis field has shown that when children are placed with others, particularly those less similarto themselves, the ability to take the perspective of another is advanced (Cassidy and others,2005; Dunn, 2004; Perner and others, 1994). The mechanism is direct, through child–childinteractions, and also mediated through the language use of facilitating adults who, incontexts that provide social challenge, are more frequently required to provide explanationsof the perspectives of others (Dunn, 2004). It would seem difference provokes social thought,social discussion and social learning.

Conclusion

Our findings provide evidence that ECEC settings do not always act as meeting places fordiverse groups. When they do, however, the quality of the relationships is highest. Thefindings direct attention to the need for further research. There is a need to identify policyand pedagogical practices that advance social connection and to investigate underlyingmechanisms that sustain these connections across time.

338 Karen Thorpe et al.

� 2010 The Author(s) CHILDREN & SOCIETY Vol. 26, 328–340 (2012)

Children & Society � 2010 National Children’s Bureau and Blackwell Publishing Limited

Page 12: Testing the Vision: Preschool Settings as Places for Meeting, Bonding and Bridging

Acknowledgements

This research was supported under Australian Research Council’s Discovery Projects fundingscheme (grant number DP0666254). We thank the Australian Twin Registry and AustralianMultiple Birth Association, Catholic Education (Queensland), Education Queensland, partici-pating members of Independent Schools (Queensland) and the Creche and Kindergarten Asso-ciation of Queensland, for assistance in recruiting children to the study. We also thank theparticipating children, families, schools and teachers who gave generously their time. Thankyou to research assistants Cathy Thompson, Maryanne Theobald, Toby Thompson and SandyHouen who conducted the child interviews with great skill. Thank you to postgraduatestudents Donna Gifford and Michael Herd who, with Rob Morgan, coded all transcripts.

Notes

1 Since the time of this study (2006–2007), childcare and early education provision inQueensland have been placed under the auspices of a single government agency.

References

Aboud FE, Mendelson MJ. 1996. Determinants of friendship selection and quality: developmental per-spectives. In The Company They Keep: Friendship in Childhood and Adolescence. Bukowskie WN,Newcombe AF, Hartup WW (eds). Cambridge University Press: New York; 87–112.

Almedom AM. 2005. Social capital and mental health: an interdisciplinary review of primary evidence.Social Science and Medicine 61: 943–964.

Ball S, Vincent C. 1998. ‘‘I heard it on the grapevine’’: ‘Hot’ knowledge and school choice. British Jour-nal of the Sociology of Education 19: 377–400.

Baum FE, Ziersch AM. 2003. Social capital. Journal of Epidemiology and Community Health 57: 320–323.Berndt TJ. 2004. Children’s friendship: shifts over half a century. Merrill-Palmer Quarterly 50: 206–223.Buysse V, Goldmand BD, Skinner ML. 2002. Setting effects on friendship formation among young chil-

dren with and without disabilities. Exceptional Children 68: 503–517.Buysse V, Goldman BD, Skinner ML. 2003a. Friendship formation in inclusive early childhood class-

rooms: what is the teacher’s role? Early Childhood Research Quarterly 18: 485–501.Buysse V, Sparkman KL, Wesley PW. 2003b. Communities of practice: connecting what we know with

what we do. Exceptional Children 69: 263–277.Cassidy KW, Fineberg DS, Brown K, Perkins A. 2005. Theory of mind may be contagious, but you can’t

catch it from your twin. Child Development 76: 97–106.Coleman JS. 1990. Foundations of Social Theory. Harvard University Press: Cambridge, MA.Council of Australian Governments. 2009. Investing in the Early Years: A National Early Childhood

Development Strategy. Australia Co (ed): Canberra.Dale PS, Dionne G, Eley TC, Plomin R. 2000. Lexical and grammatical development: a behavioral

genetic perspective. Journal of Child Language 27: 619–642.Danby S, Farrell A. 2004. Accounting for young children’s competence in educational research: new

perspectives on research ethics. Australian Educational Researcher 31: 35–49.Davies B, Davis E, Cook K, Waters E. 2007. Getting the complete picture: combining parental and child

data to identify the barriers to social inclusion for children living in low socio-economic areas. Child:Care, Health and Development 34: 214–222.

Dunn J. 2004. Children’s Friendships: The Beginnings of Intimacy. Blackwell: Malden.Farran DC, Son-Yarbrough W. 2001. Title I funded preschools as a developmental context for children’s

play and verbal behaviors. Early Childhood Research Quarterly 16: 245–262.Farrell A, Tayler C, Tennent L. 2004. Building social capital in early childhood education and care: an

Australian study. British Educational Research Journal 30: 623–632.

ECEC and Social Capital 339

� 2010 The Author(s) CHILDREN & SOCIETY Vol. 26, 328–340 (2012)

Children & Society � 2010 National Children’s Bureau and Blackwell Publishing Limited

Page 13: Testing the Vision: Preschool Settings as Places for Meeting, Bonding and Bridging

Goodman R. 1997. The Strengths and Difficulties Questionnaire: a research note. Journal of ChildPsychology and Psychiatry 38: 581–586.

Guralnick MJ, Connor RT, Hammond MA, Gottman JM, Kinnish K. 1996. The peer relations ofpreschool children with communication disorders. Child Development 67: 471–489.

Haselager GJT, Hartup WM, van Lieshout CFM, Riksen-Walraven JM. 1998. Similarities between friendsand non-friends in middle childhood. Child Development 69: 1198–1208.

Hay DA, Preedy P. 2006. Meeting the educational needs of multiple birth children. Early Human Devel-opment 82: 397–403.

Howes C, Wu F. 1990. Peer interactions and friendships in an ethnically diverse school setting. ChildDevelopment 61: 537–541.

James A, Prout A. 1997. Constructing and Reconstructing Childhood. Contemporary Issues in the Socio-logical Study of Childhood. Falmer Press: London.

Morrow V. 2001. Networks and Neighbourhood: Children’s and Young People’s Perspectives. NHS HealthDevelopment Agency: London.

Newcomb AF, Bagwell CL. 1995. Children’s friendship relations: a meta-analytic review. PsychologicalBulletin 117: 306–347.

OECD. 2006. Starting Strong II : Early Childhood Education and Care. OECD Publishing: Paris.Perner J, Ruffman T, Leekam SR. 1994. Theory of mind is contagious: you catch it from your sibs. Child

Development 65: 1228–1238.Perren S, Alsaker FD. 2006. Social behaviour and peer relationships of victims, bully-victims and

bullies in kindergarten. Journal of Child Psychology and Psychiatry 47: 45–57.Prieto HR, Sahlstrom F, Calander F, Heikkila M 2002. Together? On childcare as a meeting place in a

Swedish city. Scandinavian Journal of Educational Research 47:43–62.Reynolds A, Temple J. 2008. Cost-effective early childhood development programs from preschool to

year 3. Annual Review of Clinical Psychology 4: 109–139.Rubin KH, Bukowski WM, Parker JG. 2006. Peer interactions, relationships, and groups. In Handbook of

Child Psychology. Damon W, Lerner RM, Eisenberg N (eds). Wiley: New York; 571–645.Schweinhart L, Montie J, Xiang Z, Barnett W, Belfield C, Nores M. 2005. Lifetime Effects: The High ⁄

Scope Perry Preschool Study Through Age 40. High ⁄ Scope Press: Michigan.Siegel S, Castellan NJ Jr. 1988. Nonparametric Statistics for the Behavioral Sciences, 2nd edn. McGraw-

Hill: London.Sylva K, Melhuish EC, Sammons P, Siraj-Blatchford I, Taggard B. 2004. The Effective Provision of Pre-school

Education (EPPE) Project: Final Report. DfES ⁄ Institute of Education, University of London: London.Thorpe K. 2003a. Twins and friendship. Twin Research 6: 532–535.Thorpe K. 2006. Twin children’s language development. Early Human Development 82: 387–395.Thorpe K, Gardner K. 2006. Twins and their friendships: differences between MZ, DZ (same sex), and

DZ (mixed sex) pairs. Twin Research and Human Genetics 9: 155–164.Thorpe K, Tayler C, Bridgestock R, Grieshaber S, Skoien P, Danby S, Petriwskyj A. 2004. Preparing For

School: Report of the Queensland Preparing for School Trials 2003 ⁄ 4. Queensland University ofTechnology, Faculty of Early Childhood: Brisbane.

Thorpe K, Tayler C, Thompson C, Staton S, Bell-Booth R 2009. Communities for Children: CentacareEvaluation Report. QUT: Brisbane.

Vincent C, Ball S. 2006. Childcare, Choice and Class Practices: Middle Class Parents and their Children.Routledge: London.

*Correspondence to: Prof. Karen Thorpe, School of Psychology and Counseling, Queensland University of Techno-

logy, Victoria Park Road, Kelvin Grove, Qld 4059, Australia, Tel.: +617 3138 3200; Fax: +617 3138 4660. E-mail:

[email protected]

Accepted for publication 16 July 2010

340 Karen Thorpe et al.

� 2010 The Author(s) CHILDREN & SOCIETY Vol. 26, 328–340 (2012)

Children & Society � 2010 National Children’s Bureau and Blackwell Publishing Limited