13
Social Antecedents of Childrens Trustworthiness Ken J. Rotenberg a, * , Lucy R. Betts b , Manuel Eisner c and Denis Ribeaud d a Keele University, Psychology Department, Keele, UK b Nottingham Trent University, Psychology Department, Nottingham, UK c University of Cambridge, Institute of Criminology, Cambridge, UK d University of Zurich, Institute of Education Science, Zurich, Switzerland A total of 1329 children were tested twice across 1 year (M = 7 years 5 months of age at Time 1 (T1)) in the Zurich Project on Social Development. The measures at T1 were corporal punishment, neighbourhood trustworthiness and childrens trustworthiness (not lying/cheating and not stealing). At Time 2 (T2), children reported the promise keeping of their classmates, which, via social relations analyses, yielded evidence for individual differences in reliability trustworthiness. Structural equation modelling analyses conrmed that there was stability in childrens trustworthiness as a latent variable. The structural equation modelling further yielded evidence that (1) corporal punishment at T1 was negatively associated with childrens trustworthiness at T1 and negatively predicted changes in childrens trustworthiness and (2) neighbour- hood trustworthiness at T1 was positively associated with childrens trustworthiness at T1 and positively predicted changes in childrens trustworthiness. The ndings yielded support for the hypotheses that corporal punishment negatively, and neighbourhood trustworthiness positively, contributes to the development of trustworthiness in children. Copyright © 2011 John Wiley & Sons, Ltd. Key words: trustworthiness; children; social antecedents Trustworthiness (e.g. lying, stealing and failing to full promises) has been the focus of investigation from the beginning of contemporary psychology (Hartshorne & May, 1928; Hartshorne, May, & Maller, 1929; Hartshorne, May & Shuttlesworth, 1930), and it continues to be so until the present (e.g. Stouthamer-Loeber, 1986; Talwar & Lee, 2008; Xu, Bao, Fu, Talwar, & Lee, 2010). Researchers have found that trustworthiness during childhood or adolescence is associated with pro- social behaviour (Rotenberg, Fox et al., 2005), sophistication of theory of mind (Talwar, Gordon, & Lee, 2007), psychological adjustment (Gervais, Tremblay, * Correspondence to: Ken J. Rotenberg, Psychology Department, Keele University, Newcastle- under-Lyme, Keele ST5 5BG, UK. E-mail: [email protected] Infant and Child Development Inf. Child Dev. 21: 310322 (2012) Published online 27 September 2011 in Wiley Online Library (wileyonlinelibrary.com). DOI: 10.1002/icd.751 Copyright © 2011 John Wiley & Sons, Ltd.

Social Antecedents of Children's Trustworthiness

Embed Size (px)

Citation preview

Infant and Child DevelopmentInf. Child Dev. 21: 310–322 (2012)Published online 27 September 2011 in Wiley Online Library(wileyonlinelibrary.com). DOI: 10.1002/icd.751

Social Antecedents of Children’sTrustworthiness

* Correspondenunder-Lyme, K

Copyright © 201

Ken J. Rotenberga,*, Lucy R. Bettsb, Manuel Eisnerc

and Denis Ribeaudd

aKeele University, Psychology Department, Keele, UKbNottingham Trent University, Psychology Department, Nottingham, UKcUniversity of Cambridge, Institute of Criminology, Cambridge, UKdUniversity of Zurich, Institute of Education Science, Zurich, Switzerland

A total of 1329 children were tested twice across 1 year (M= 7 years5months of age at Time 1 (T1)) in the Zurich Project on SocialDevelopment. The measures at T1 were corporal punishment,neighbourhood trustworthiness and children’s trustworthiness(not lying/cheating and not stealing). At Time 2 (T2), childrenreported the promise keeping of their classmates, which, via socialrelations analyses, yielded evidence for individual differences inreliability trustworthiness. Structural equationmodelling analysesconfirmed that there was stability in children’s trustworthiness as alatent variable. The structural equation modelling further yieldedevidence that (1) corporal punishment at T1 was negativelyassociated with children’s trustworthiness at T1 and negativelypredicted changes in children’s trustworthiness and (2) neighbour-hood trustworthiness at T1 was positively associated with children’strustworthiness at T1 and positively predicted changes in children’strustworthiness. The findings yielded support for the hypotheses thatcorporal punishment negatively, and neighbourhood trustworthinesspositively, contributes to the development of trustworthiness inchildren. Copyright © 2011 John Wiley & Sons, Ltd.

Key words: trustworthiness; children; social antecedents

Trustworthiness (e.g. lying, stealing and failing to fulfil promises) has been the focusof investigation from the beginning of contemporary psychology (Hartshorne &May, 1928; Hartshorne, May, & Maller, 1929; Hartshorne, May & Shuttlesworth,1930), and it continues to be so until the present (e.g. Stouthamer-Loeber, 1986;Talwar & Lee, 2008; Xu, Bao, Fu, Talwar, & Lee, 2010). Researchers have foundthat trustworthiness during childhood or adolescence is associated with pro-social behaviour (Rotenberg, Fox et al., 2005), sophistication of theory of mind(Talwar, Gordon, & Lee, 2007), psychological adjustment (Gervais, Tremblay,

ce to: Ken J. Rotenberg, Psychology Department, Keele University, Newcastle-eele ST5 5BG, UK. E-mail: [email protected]

1 John Wiley & Sons, Ltd.

Social Antecedents of Trustworthiness 311

Desmarais-Gervais, & Vitaro, 2000; Stouthamer-Loeber, 1986), academic achieve-ment (Imber, 1973), number of friendships (Rotenberg, Michalik, Eisenberg, & Betts,2004), reflectivity (Rotenberg et al., 2008) and self-esteem (Betts & Rotenberg, 2007).It has been found that girls demonstrate greater trustworthiness (e.g. keepingpromises) than do boys (Rotenberg et al. 2004).

Despite the breadth of research on trustworthiness during childhood andadolescence, there is a dearth of research on the psychosocial factors that contributeto its development. The current research was carried out to redress that oversightby examining the extent to which neighbourhood trustworthiness and corporalpunishment are concurrently associated with, and longitudinally predict, children’strustworthiness.

Conceptualization of Trustworthiness

Trustworthiness was conceptualized in the current research according to Rotenbergand his colleagues’ 3 Bases� 3 Domains� 2 Target Dimensions (BDT) frameworkof interpersonal trust (Rotenberg, 1994, 2010). The framework includes the followingthree bases of trust: (1) reliability, which refers to a person fulfilling his or her wordand promise; (2) emotional trust, which refers to a person refraining from causingemotional harm (e.g. keeping secrets); and (3) honesty, which refers to a persontelling the truth and engaging behaviours that are guided by benign rather than ma-licious intent and genuine rather thanmanipulative strategies. The three domains (ascrossed with the bases) are (1) cognitive/affect, which comprises individuals’ beliefsand feelings that others demonstrate the three bases of trust; (2) behaviourdependent, which comprises individuals behaviourally relying on others to act in atrusting fashion as per the three bases of trust; and (3) behaviour enacting (i.e. trust-worthiness), which comprises individuals behaviourally engaging in the three bases oftrust. The current study examined the reliability and honesty bases of trustworthiness.

Neighbourhood Trustworthiness as an Antecedent of Children’s Trustworthiness

Neighbourhood trustworthiness has been found to be associated with children’spsychosocial functioning. In particular, Meltzer, Vostanis, Goodman, and Ford(2007) found that 11- to 16-year-old children’s perception of their neighbourhoodsas trustworthy and honest was associated with their lack of psychopathology,notably low levels of emotional disorders. There are a number of reasons to expectthat trustworthiness of neighbourhood is associatedwith children’s trustworthiness.The relation may be found because (1) neighbours serve as models for children(Rotenberg, 1995); (2) a trustworthy neighbourhood promotes trustworthiness inchildren by dyadic interactions between trustworthy community members and chil-dren (Rotenberg, 2010); and (3) the trustworthiness of the neighbourhood is a sourceof social capital (a value of the social network) that engenders trustworthiness inchildren (Meltzer et al., 2007). The current study was designed to examine thehypothesis that neighbourhood trustworthiness positively contributes to trust-worthiness in children. It was specifically expected that neighbourhood trustworthi-ness would be positively associated with, and would positively predict changes in,children’s trustworthiness.

Parenting Style as an Antecedent of Children’s Trustworthiness

There is a substantive body of research that has examined the effects of parentingpractices on children’s social behaviour (see Kaisa & Nurmi, 2005). Guided by

Copyright © 2011 John Wiley & Sons, Ltd. Inf. Child Dev. 21: 310–322 (2012)DOI: 10.1002/icd

312 K. J. Rotenberg et al.

moral and theoretical issues, several researchers have focused on the effects ofcorporal punishment on children’s social functioning (see Gershoff, 2002). Bothnarrative reports (e.g. Larzelere, 2000) and meta-analyses (e.g. Gershoff, 2002)yield support for the conclusion that corporal punishment is associated with anti-social behaviour (e.g. cheating, lying, hostility, and emotional problems) anddelinquent behaviour (e.g. self-reported truancy and underage drinking).There are few longitudinal studies of the relations between corporal punishmentand antisocial/delinquent behaviour (see Gershoff, 2002). Recent longitudinalstudies employing growth curve analyses have shown, though, that corporalpunishment places children at risk for the development of such behaviour (Berlinet al., 2009; Grogan-Kaylor, 2005; Lansford, Criss, Dodge, Shaw, Pettit, & Bates,2009). The research also has yielded gender differences: boys are more likely tobe the target of corporal punishment than are girls (see Berlin et al., 2009). Further-more, the meta-analysis by Gershoff (2002) showed that gender moderated the re-lation between corporal punishment and problem behaviour: it was found thatthe more girls were included in the study, the less corporal punishment was asso-ciated with aggression and antisocial behaviour.

There are two limitations with the research as it pertains to the questionaddressed by the current study. First, with an exception, previous research hasexamined broadly based measures of antisocial behaviour or delinquent behaviourand therefore it is unclear whether corporal punishment is negatively linked totrustworthiness per se. It should be emphasized that, although trustworthiness is(in the negative form of untrustworthiness) a form of antisocial or delinquentbehaviour, it is a conceptually unique social domain that includes reliability,emotional trustworthiness and honesty (Rotenberg, 2010). Second, research hasnot adequately examined the longitudinal relation between corporal punishmentand trustworthiness which would yield evidence for probable causation.

As the exception noted, Talwar and Lee (in press) tested 3- to 4-year-old chil-dren from West Africa and found that children from punitive schools were morelikely to lie about yielding to temptation (i.e. peeking at a toy) and were better ableto maintain their deception than were children from non-punitive schools. Accord-ing to the researchers, the findings support the conclusion that a punitive environ-ment fosters children’s dishonesty and their ability to lie to conceal theirtransgressions. As one account of their findings, Talwar and Lee (in press)argued that children from punitive schools were more likely to learn to lieand to successfully deceive about lying than were children from non-punitiveschools to avoid the negative consequences of punishment.

It is worthwhile to consider the preceding study from an attribution theoryperspective. On the basis of the attribution theory, researchers (e.g. Dix & Grusec,1985; Hoffman, 1983; Lepper, 1983) have proposed that parents’ use of physicalpunishment undermines the development of internalization of moral principlesin children. For example, Lepper (1973) found that children who had receivedsevere threat from an adult prohibiting them from playing with a desired toy sub-sequently showed less resistance to temptation and lower self-attributions ofhonesty than children in a control group. Lepper (1973) hypothesized that childrenwho had been the recipients of severe threat experienced overjustification,whereby they attributed their resistance to temptation to the threat, inferred theywere not honest and subsequently engaged in less honest behaviour. On the basisof the attribution theory, it has been argued that although parents may use punish-ment to ensure trustworthiness in their children, it has the opposite effect becauseit provides a salient justification for the children’s compliance (Dix & Grusec, 1985;Hoffman, 1983; Lepper, 1973).

Copyright © 2011 John Wiley & Sons, Ltd. Inf. Child Dev. 21: 310–322 (2012)DOI: 10.1002/icd

Social Antecedents of Trustworthiness 313

The social learning and attribution principles may serve as complementaryprocesses that contribute to how a punitive environment (including corporalpunishment) affects children’s trustworthiness, as discussed in the study of Talwarand Lee (in press). Punitive schools may promote lying and concealment of lyingwhen the children transgressed (i.e. peeked) because they were motivated to avoidpunishment. According to the attribution theory though, children from punitiveschools who did not transgress also were at risk for dishonesty: specifically, theywould have attributed their compliance (i.e. not peeking) to the punitive practicesof the school rather than to their own personal honesty and thus are less inclined tobehave honestly in the future. Regardless of behaviour then, children from puni-tive schools would have failed to establish internalized trustworthiness eitherbecause of dishonest behaviour (and corresponding self-perception) or becauseof the lack of self-attribution of honesty when complying. The complementaryeffects of learning and attribution principles lead to the expectation that corporalpunishment, as a punitive practice, would be negatively linked to trustworthinessin children.

One question of interest is when in the course of development would corporalpunishment be negatively linked to trustworthiness in children? The focus of thecurrent study was on trustworthiness as an internalized characteristic comprisinga self-ascribed standard accompanied by the experience of guilt when violated.There is an extensive body of research that has examined the internalization ofmorality during childhood. Kochanska and colleagues (e.g. Kochanska & Aksan,2006; Kochanska, Gross, Lin, & Nichols, 2002) have examined conscience duringearly childhood and found, for example, that it is negatively predicted by powerassertive parenting (including physical punishment). Other researchers haveexamined the development of internalization (including conscience) from earlychildhood to adolescence from the framework of moral–emotional responsiveness(Stilwell, Galvin, Kopta, & Norton, 1994) and moral reasoning in accordance withthe Kohlbergian theory (Dawson, 2002). The research may be taken to suggest thatthe internalization of morality is a multifaceted phenomenon that emerges fromearly childhood to adolescence. Nevertheless, reviews of the literature (seeThompson et al., 2006; Smetana, 2006) converge in suggesting that the onset ofmiddle childhood (approximately 7 to 8 years of age) is a significant period forthe formation of the internalization of morality (e.g. children’s comprehensionof guilt) and, thus, likely internalized trustworthiness. For that reason, the currentstudy was designed to investigate the extent to which corporal punishment as aparenting style is negatively linked to trustworthiness in children across the 7-to 8-year span.

Overview of the Current Investigation

The current investigation comprised a 1-year longitudinal design that tested thehypothesis that neighbourhood trustworthiness would be concurrently associatedwith children’s trustworthiness and positively predict changes in children’s trust-worthiness. Also, it was hypothesized that corporal punishment would be concur-rently (and negatively) associated with children’s trustworthiness and negativelypredict changes in children’s trustworthiness. Finally, gender differences wereexpected: it was anticipated that higher corporal punishment would be receivedby boys than girls (see Berlin et al., 2009) and higher trustworthiness would beshown by girls than by boys (see Rotenberg et al., 2004). The study was furtherdesigned to (1) explore whether gender moderated the hypothesized relationsoverall and (2) specifically whether gender moderated the hypothesized relations

Copyright © 2011 John Wiley & Sons, Ltd. Inf. Child Dev. 21: 310–322 (2012)DOI: 10.1002/icd

314 K. J. Rotenberg et al.

between corporal punishment and trustworthiness because the relations betweencorporal punishment and problem behaviour have been found to be moderatedby gender (see Gershoff, 2002).

METHOD

Participants

The data for this study were derived from the Zurich Project on the Social Devel-opment of Children (z-proso), which is an ongoing prospective longitudinal studyof a cohort of children who began primary school in the year 2004 (i.e. average ageof 7) (for a more detailed overview, see Eisner & Ribeaud, 2007). The sample sizeswere 1355 children at Time 1 (T1) and 1329 at Time 2 (T2), but the sample sizesvaried by measure. The mean age of the children was 7 years 5months (SD= 4.6months) at T1 and 8 years 1month (SD= 4.6months) at T2. The sampling framewas formed by all the 90 public primary schools in the City of Zurich. Schoolswere first blocked by school size and socio-economic background of the schooldistrict, and then a stratified sample of 56 schools was drawn. A total of 1329 parentsagreed to participate in the study at T1.

Measures

Child trustworthiness (teacher-reported)The honesty basis of trustworthiness was assessed by two items from the Non-

aggressive Conduct Disorder scale that is part of the Social Behaviour Question-naire (Tremblay et al., 1991). The Non-aggressive Conduct Disorder scale has beenused extensively to assess antisocial behaviour in children (e.g. Stemmler, Lösel,Beelmann, Jaursch, & Zenkert, 2005). The two items were (1) ‘the child lies andcheats’ (lying and cheating) and (2) ‘the child steals at school’ (stealing). The itemswere rated by teachers on a 5-point Likert scale ranging from 1=never to 5= veryoften at T1 and T2. The direction of the items was reversed such that the greaterscores corresponded to not lying/cheating and not stealing and, thus, to greatertrustworthiness.

Child trustworthiness (peer reported)The reliability basis of trustworthiness was assessed by the procedure and

analyses developed by Rotenberg and his colleagues (Betts & Rotenberg, 2008;Rotenberg, MacDonald, & King, 2004). This measure was assessed at T2 onlybecause of progressive changes in researchers involved in the z-proso project.The participants were read a list of their classmates by the researcher and askedto rate each one on a 5-point scale range from 1= never to 5 = always as to ‘howoften each classmate keeps promises he/she has made’. The researcher told parti-cipants that keeping a promise is ‘when someone said he or she would do some-thing and did it’. The Likert judgments were subjected to a Social RelationAnalysis (see Betts & Rotenberg, 2008; Betts, Rotenberg, & Trueman, 2010).

The Social Relations Model (Kenny & La Voie, 1984) examines variance withindyadic interactions and permits the investigation to assess how much of thebehaviour is due to the individual dyad members and how much of the behaviouris due to the unique relationship between them (Malloy & Kenny, 1986). Whenapplied to trust, the Social Relations Model examines the extent to which individualstrust their interaction partners regardless of who they interact with (termed actor

Copyright © 2011 John Wiley & Sons, Ltd. Inf. Child Dev. 21: 310–322 (2012)DOI: 10.1002/icd

Social Antecedents of Trustworthiness 315

variance and actor effects), the extent to which individuals elicit ratings of trust fromtheir interaction partners regardless of who they interact with (termed partnervariance and partner effects) and the extent to which individuals trust their inter-action partner because of the unique relationship between them (termed relationshipvariance, see Betts & Rotenberg, 2008, and Betts et al., 2010). The social relationsanalysis yielded significant partner variance, .145, p< .05. The reliability of this effectestimate was .670 based on a group size of 10.85. The findings confirm that therewere significant individual differences in the participants’ reliability trustworthiness(promise keeping) as reported by peers. Larger scores denoted that the participantsshow greater reliability trustworthiness as reported by peers.

Child trustworthiness scale scores (across informants at T2)The scores on the ‘not lying and cheating’ and ‘not stealing’ at T1were appreciably

correlated (r(1325) = .39, p< .001) and, when transformed to z-scores, were summedto yield a Child’s Trustworthiness scale score for T1 in the correlational analyses.The scores on the ‘not lying and cheating’ at T2, ‘not stealing’ at T2 and promisekeeping at T2 were transformed to z-scores and then summed to yield a Child’sTrustworthiness scale score for T2 in the correlational analyses (a= .64, as evidence foracceptable internal consistency). Larger scores denoted greater Child Trustworthinessscale scores at T1 and T2.

Corporal punishment (parent reported)This was assessed by the Corporal Punishment subscale of the Alabama Parent-

ing Questionnaire (Shelton, Frick, & Wootton, 1996). This is a three-item scalecomprising the following items: (1) ‘You spank your child with your hand whenshe/he has done something wrong’; (2) ‘You slap your child when he/she has donesomething wrong’; and (3) ‘You hit your child with a belt, switch, or other objectwhen he/she has done something wrong’. Parents rated on a 5-point frequencyscale ranging from 1= never to 5 = always. This scale demonstrated modest internalconsistency, a= .54, but the number of items was modest. This level of internalconsistency of the Corporal Punishment subscale (or its factor score) of the AlabamaParenting Questionnaire (a= .50 s) has been found in the majority of studies (Clerkin,Marks, Policaro, &Halperin, 2007; Dadds,Maujean, & Fraser, 2003; Hawes &Dadds,2006), although lower and higher internal consistency have been found in individualstudies (Essau Sasagawa, & Frick, 2006; Shelton et al., 1996). Researchers have foundacceptable test–retest reliability of the Corporal Punishment subscale across a 2-weekspan, r= .90 (Dadds et al., 2003). Also, research has yielded evidence for the validityof the Corporal Punishment scale by factor analyses (Clerkin et al., 2007; Essau et al.,2006) and by the association between that scale and aversive parent behaviour inparent–child interactions (Hawes & Dadds, 2006).

Neighbourhood trustworthiness (parent reported)Thiswas assessed at T1 by selected items from the Social Network scale developed

by Oberwittler (2007). This comprised four items: (1) ‘People in this neighbourhoodcan be trusted’; (2) ‘Watched your neighbors’ property when they are out of town’;(3) ‘Talked to neighbors about personal things’; and (4) ‘Taken care of neighbours’children while the neighbors were away from home’. The first item was rated on a4-point scale, from 1= often, 2 = sometimes, 3 = rarely and 4=never, and theremaining items were rated for the frequency with which the individual engagedin the activity. The ratings were reversed and thus summed such that higher scores

Copyright © 2011 John Wiley & Sons, Ltd. Inf. Child Dev. 21: 310–322 (2012)DOI: 10.1002/icd

316 K. J. Rotenberg et al.

corresponded to higher trustworthiness of the neighbourhood. The items weresubjected to a preliminary factor analysis (principal components) with a varimaxrotation. The factor analysis yielded one factor with an eigenvalue=2.12 thataccounted for 53% of the variance with loadings of 58, 81, .77 and .73 on items 1 to4, respectively. The items were summed to create a neighbourhood trustworthinessscale at T1 in the correlational analyses. There was acceptable internal consistencyof the neighbourhood trustworthiness scale, a= .70.

Procedure

The current data came from the first (T1) and second (T2) waves of data collectionof the Zurich Project on the Social Development of Children (Eisner & Ribeaud,2007) when the children were at the age of 7 (Time 1, M= 7 years 5months, SD=4.6months) and then again at the age of 8 (M= 8 years 1month, SD= 4.6months).Each sweep (Time) comprises data collection from the parent, the child and theteacher. The testing comprised computer-assisted face-to-face parent interviews(parent), interviews with the children primarily held in the schools and teacherassessments involving paper-and-pencil questionnaires.

The data utilized in the current study comprised the following: (1) the measuresof the children’s honesty trustworthiness at T1 and T2 as reported by teachers;(2) children’s reliability trustworthiness at T2 as reported by peers and derivedfrom the social relations analyses; (3) neighbourhood trustworthiness at T1 asreported by parents; and (4) corporal punishment at T1 as reported by parents.The measures were drawn from reports by varied informants to minimize effectsof common method variance on associations among the measures in the design.

Results

Correlations among the measuresThe correlations among the measures (with means and standard deviations) are

shown in Table 1. Corporal punishment at T1 was negatively correlated withneighbourhood trustworthiness at T1, and at T2, and child trustworthiness at T1and T2. Neighbourhood trustworthiness at T1 was correlated with Child Trust-worthiness at T1 and T2. Child trustworthiness at T1 was correlated with childtrustworthiness at T2.

Gender differencesThe measures were subjected to a multivariate analysis with gender as the

between factor. The multivariate analysis of variance yielded significant effects

Table 1. Correlations among the measures with means and standard deviations (SDs)

Measure M SD NT1 CTW1 CTW2

Corporal punishment at T1 (CP1) .45 .49 �.19*** �.08** �.19***Neighbourhood trustworthiness at T1 (NT1) 11.81 3.18 .14*** .18***Child trustworthiness at T1 (CTW1) 13.54 1.67 .57***Child trustworthiness at T2 (CTW2) 12.27 2.22

**p< .01 and***p< .001; the dfs ranged from 837 to 1085.

Copyright © 2011 John Wiley & Sons, Ltd. Inf. Child Dev. 21: 310–322 (2012)DOI: 10.1002/icd

Social Antecedents of Trustworthiness 317

of gender, F(4, 811) = 3.34, p< .016, �2 = .016 (Pillai’s Trace). Analyses of varianceshowed that there were gender differences on child trustworthiness at T2,F(1, 814)=3.93, p< .048, �2= .005, and corporal punishment at T1, F(1, 814)=8.87,p= .003, �2= .011. The findings confirmed that (1) girls demonstrated higher childtrustworthiness than did boys at T2, M=12.51 (SD=2.28) and M=12.23 (SD=2.14),respectively, and (2) boys received greater corporal punishment at T1 than did girls,M= .51 (SD= .51) and M= .41 (SD= .46), respectively.

Structural equation modellingThe model tested is shown in Figure 1. The structural equation modelling

analysis of Model 1 showed that it was a good fit of the data. The Normed FitIndex= .94, Comparative Fit Index= .95 and Root Mean Square Error of Approxima-tion= .044, although the w2(29)= 123.13, p< .001, is statistically significant. Researchershave regarded Comparative Fit Index> .90 and Root Mean Square Error of Approxi-mation< .060 as a good fit (Bentler, 1990). Although a non-significant w2 is regarded asfulfilling good fit, it has been recognized that this requirement is conventionallyviolated when samples are large (see Bentler, 1990) as was the case in the currentresearch. All paths and covariances in the model attained statistical significance(ps< .05). As evidence for measure stability, child trustworthiness at T1 positivelypredicted Child Trustworthiness at T2. Consistent with the hypothesis, it was found

NeighbourhoodTrustworthiness

at T1

Disclosureto Neighbours

at T1E4

.53

WatchedNeighbourhoodChildren at T1

E3

.60

TrustNeighbours

at T1E2 .68

Taken Careof Neighbours'

Childrenat T1

E1

CorporalPunishment

at T1

.47-.13

.32

ChildTrustworthiness

at T1

Not Lying& Cheating

at T1E6

.83

Not Stealingat T1

E5 .47Child

Trustworthinessat T2

Not Stealingat T2

E7

.43

Not Lying& Cheating at T2 E8

KeepingPromises

at T2E9

.42

-.07

.11

.25

.64

.19

-.12

.96

Figure 1. Structural equation modelling of the relations among the measures. All paths andcovariances in the model attained statistical significance at p< .05.

Copyright © 2011 John Wiley & Sons, Ltd. Inf. Child Dev. 21: 310–322 (2012)DOI: 10.1002/icd

318 K. J. Rotenberg et al.

that (1) there was a negative covariance between corporal punishment at T1 and childtrustworthiness at T1; (2) corporal punishment discipline at T1 negatively predictedchanges in child trustworthiness. Also, consistent with the hypothesis, it was foundthat (1) there was a positive covariance between neighbourhood trustworthiness atT1 and child trustworthiness at T1 and (2) neighbourhood trustworthiness at T1positively predicted changes in child trustworthiness.

The issue of whether gender moderated the relations in the model was examinedusing the two statistical procedures. First, to test whether gender moderated allrelations in the model (factors, covariances and paths), we carried out an omnibustest (see Byrne, 2001) that compared the multigroup baseline model with the modelcomprising equality constraints by gender. The difference in the models wasw2(10) = 12.73, p= .22, which failed to attain significance and thus demonstrated thatgender did not moderate the relations in the model. Second, a specific test wascarried out (see Byrne, 2001) that compared the multigroup baseline model withthe following model constrained by gender: (1) the covariance between corporalpunishment at T1 and child trustworthiness at T1 and (2) the path between corporalpunishment at T1 and child trustworthiness at T2. The contrast yielded w2(2) = 5.3,p= .07, which did not attain significance but did approach significance. This trendappeared to be because of the path between the corporal punishment at T1 andchild trustworthiness at T2 being evident for boys, b=�.04, p= .02, but not for girls,b= .01.

DISCUSSIONThe findings yielded support for the hypothesized relations between children’strustworthiness and both corporal punishment and neighbourhood trustworthiness.As hypothesized, it was found that corporal punishment was negatively associatedwith children’s trustworthiness and negatively predicted changes in children’s trust-worthiness. Also, as hypothesized, it was found that neighbourhood trustworthinesswas positively associated with children’s trustworthiness and positively predictedchanges in children’s trustworthiness. In addition, it was found that corporalpunishment was concurrently negatively associated with neighbourhood trust-worthiness. Finally, consistent with the hypotheses, higher corporal punishmentwas received by boys than girls, and girls demonstrated higher trustworthinessthan did boys at Time 2. Gender was not found to moderate the relations eitherfor the model in total or for the hypothesized relations with corporal punishment.

The findings regarding neighbourhood trustworthiness complement researchcarried out by Meltzer et al. (2007), which showed that children’s perception oftheir neighbourhoods as trustworthy and honest was associated with their lackof psychopathology, notably low levels of emotional disorders. The current studyuniquely shows that neighbourhood trustworthiness was concurrently associatedwith children’s trustworthiness and positively predicted changes in children’strustworthiness. The latter longitudinal relation is consistent with the conclusionthat neighbourhood trustworthiness is a probable cause of, and thus contributesto, trustworthiness in children.

The potential mechanisms responsible for the aforementioned relations are thefollowing: (1) neighbours serve as models for children (Rotenberg, 1995); (2) atrustworthy neighbourhood promotes trustworthiness in children by dyadicprocesses (Rotenberg, 2010); and (3) trustworthy neighbourhood are a source ofsocial capital that engenders trustworthiness in children (Meltzer et al., 2007).If the observed longitudinal relation between neighbourhood and children’s

Copyright © 2011 John Wiley & Sons, Ltd. Inf. Child Dev. 21: 310–322 (2012)DOI: 10.1002/icd

Social Antecedents of Trustworthiness 319

trustworthiness is the result of dyadic processes, then the relation may be a product ofcomplex interdependent interactions between children and their neighbours (i.e.exchanges of trustworthy behaviour). In future, researchers should examine the extentto which those mechanisms are responsible for the observed relation.

The findings yielded by the current study complement a range of studiesdemonstrating that corporal punishment is concurrently associated with antisocialand delinquent behaviour (see Gershoff, 2002; Larzelere, 2000) and with longitu-dinal studies demonstrating that corporal punishment places children at risk forthe development of antisocial behaviour (Berlin et al., 2009; Grogan-Kaylor, 2005;Lansford et al., 2009). The findings from the current study are unique in that theydemonstrate (1) a negative link between that corporal punishment and children’strustworthiness per se and (2) a negative longitudinal relation between corporalpunishment and children’s trustworthiness. The findings are consistent with thehypothesis that corporal punishment negatively contributes to trustworthiness inchildren. It should be emphasized that although the longitudinal findings yieldedby the current study are indicative of causal relations, the data are correlational innature and further research is required to provide definite evidence for causality.

Meta-analyses have demonstrated that the relations between corporal punish-ment and problem behaviour (antisocial and aggression) are moderated by gender(Gershoff, 2002). In the current study, gender was not found to serve as amoderator of the hypothesized relations between corporal punishment andtrustworthiness. Nevertheless, the moderating effects of gender were found toapproach significance and that appeared to be due to corporal punishmentlongitudinally (and negatively) predicting trustworthiness for boys but not forgirls. It is possible that corporal punishment has longer-term negative conse-quences for boys only, but researchers need to examine whether or not thoseeffects are replicable before any conclusions are warranted. These findings areconsistent with the meta-analysis by Gershoff (2002), which indicates that theassociation between corporal punishment and aggression/antisocial behaviouris primarily found for boys. It is worthwhile to note that in the current study,gender was not found to moderate the covariances and paths for the completemodel and, therefore, gender was not a substantive moderator of the observedrelations.

The current study was guided by the notion that corporal punishment isnegatively linked to trustworthiness in children because of learning and attributiontheory principles. According to attribution theory (e.g. Dix & Grusec, 1985;Hoffman, 1983; Lepper, 1983), corporal punishment is negatively linked to trust-worthiness in children because it undermines their self-attribution of trust-worthiness and trustworthiness when they engage in behaviour prescribed byauthorities. The study also was guided by the learning account of Talwar andLee (in press) of the effects of punitive environment on children’s social behav-iour. In the current study, children of parents who use corporal punishmentmay have acquired dishonest behaviour (notably lying) because they were (1)low in self-attributed trustworthiness and thus trustworthiness and (2) moti-vated to avoid negative consequences of corporal punishment.

The observation that corporal punishment was negatively associated with, andlongitudinally predicted, trustworthiness during the 7- to 8-year age period inthe current study is consistent with the premise that corporal punishment islinked to trustworthiness as an internalized characteristic. In that vein, researchersshould examine the extent to which self-ascribed standards of trustworthinessaccompanied by the experience of guilt when violated (as evidence for internalizedtrustworthiness) are responsible for the relations observed in the current study.

Copyright © 2011 John Wiley & Sons, Ltd. Inf. Child Dev. 21: 310–322 (2012)DOI: 10.1002/icd

320 K. J. Rotenberg et al.

Also, it would be worthwhile for researchers to examine how trustworthiness isrelated to the broader array of antisocial/delinquent behaviours and whether therelation between corporal punishment and trustworthiness observed in the currentstudy (and the processes underlying those) are responsible for the relationsbetween corporal punishment and those broader arrays of behaviour.

The current study yielded a negative association between corporal punishmentand neighbourhood trustworthiness. These are correlational findings, and there-fore, any interpretation must be regarded as speculative. It is possible, though, thatneighbourhood trustworthiness decreases the likelihood of parents engaging incorporal punishment because trustworthy neighbours diminish the need forparental supervision (e.g. they help ensure moral behaviour in children) and thusdecreases the likelihood of physical discipline. This interpretation is consistentwith research demonstrating positive relations between desirable structural orqualitative aspects of neighbourhood and parenting such as parental coachingand (low) parental solicitation (Caughy, Nettles, & O’Campo, 2007; Lima, Caughy,Nettles, & O’Campo, 2010; Vieno et al., 2010). This hypothesis warrants furtherinvestigation.

The findings yielded by the current study yield support for the utility of the basis,domain and target interpersonal trust framework (Rotenberg, 2010). Consistent withthe BDT framework, it was found that measures of honesty and reliability wereuseful indices of a latent construct of children’s trustworthiness. This pattern wasfound even though the measures were obtained from different reporting sources(i.e. teachers versus peers) and differing types of measures (i.e. raw means versusscores yielded by the social relations analysis).

Summary

Trustworthiness has been a focus of interest from the very beginning of contemporarypsychology (Hartshorne & May, 1928) and continues to be a topic of considerableinterest in psychology (e.g. Xu et al., 2010). The current study provides evidenceregarding the correlates and potential antecedents of trustworthiness (i.e. corporalpunishment and neighbourhood trustworthiness), and the findings should serve asan impetus for researchers to examine a wider range of correlates, and antecedents,as well as consequences, of children’s trustworthiness in the future.

REFERENCES

Bentler, P. M. (1990). Comparative fit indexes in structural models. Psychological Bulletin, 107,238–246.

Berlin, L. J., Ispa, J. M., Fine, M. A., Malone, P. S., Brooks-Gunn, J., Bradly-Smith, C., ... Bai, Y.(2009). Correlates and consequences of spanking and verbal punishment for low-incomeWhite, African-American, and Mexican-American toddlers. Child Development, 80,1403–1420.

Betts, L. R., & Rotenberg, K. J. (2007). Trustworthiness, friendship and self-control: Factorsthat contribute to young children’s school adjustment. Infant and Child Development, 16,491–508.

Betts, L. R. & Rotenberg, K. J. (2008). A social relations analysis of young children’s trust intheir peers across the early years of school. Social Development, 17, 1039–1055.

Betts, L. R., Rotenberg, K. J., & Trueman, M. (2010). The application of dyadic level analysesto examine young children’s specific trust beliefs, peer-reported trustworthiness andreciprocity of trust, in K. J. Rotenberg (Ed.), Interpersonal trust during childhood andadolescence (110–132). Cambridge: Cambridge University Press.

Copyright © 2011 John Wiley & Sons, Ltd. Inf. Child Dev. 21: 310–322 (2012)DOI: 10.1002/icd

Social Antecedents of Trustworthiness 321

Caughy, M., Nettles, S., & O’Campo, P. (2007). Community influences on adjustment in firstgrade: An examination of an integrated process model. Journal of Child & Family Studies,16, 819–836.

Clerkin, S. M., Marks, D. J., Policaro, K. L., Halperin, J. M. (2007). Psychometric properties ofthe Alabama Parenting Questionnaire—preschool revision. Journal of Clinical Child andAdolescent Psychology, 36, 19–28.

Dadds, M. R., Maujean, A., & Fraser, J. A. (2003). Parenting and conduct problems in children:Australian data and psychometric properties of the Alabama Parenting Questionnaire.Australian Psychologist, 38, 238–241.

Dawson, T. L. (2002). New tools, new insights: Kohlberg’s moral judgement stages revisited.International Journal of Behavioral Development, 26, 154–166.

Dix, T., & Grusec, J. E. (1985). Parent attribution processes in the socialization of children. InI. E. Sigel (Ed.), Parental belief systems (pp. 201–233). Hillsdale N J: Erlbaum.

Eisner, M., & Ribeaud, D. (2007). Conducting a criminological survey in a culturally diversecontext: Lessons from the Zurich Project on the Social Development of Children. EuropeanJournal of Criminology, 4, 271–298.

Essau, C. A. Sasagawa, S., & Frick, P. J. (2006). Psychometric properties of the AlabamaParenting Questionnaire. Journal of Child and Family Studies, 15, 597–616.

Gershoff, E. T. (2002). Corporal punishment by parents and associated child behaviors andexperiences: A meta-analytic and theoretical review. Psychological Bulletin, 128, 539–579.

Gervais, J., Tremblay, R. E., Desmarais-Gervais, L., & Vitaro, F. (2000). Children’s persistentlying, gender differences, and disruptive behaviors: A longitudinal perspective. InternationalJournal of Behavioral Development, 24, 213–221.

Grogan-Kaylor, A. (2005). Corporal punishment and the growth trajectory of children’santisocial behavior. Child Maltreatment, 10, 283–292.

Hartshorne, H., & May, M. A. (1928). Studies in the nature of character. Vol. I: Studies in deceit.New York: Macmillan.

Hartshorne, H., May, M. A., & Maller, J. B. (1929). Studies in the nature of character. Vol. II:Studies in self-control. New York: Macmillan.

Hartshorne, H., May, M. A., & Shuttleworth, F. K. (1930). Studies in the organization of character.Vol. III: Studies in the organization of character. New York: Macmillan.

Hawes, D. J., & Dadds, M. R. (2006). Assessing parenting practices through parent-report anddirect observation during parent-training. Journal of Child and Family Studies, 15, 555–568.

Hoffman, M. L. (1983). Affective and cognitive processes in moral internalization. In E. T.Higgins, D. N. Ruble, & W. W. Hartup (Eds.) Social cognition and social development(pp. 236–274). New York: Cambridge University Press.

Imber, S. (1973). Relationship of trust to academic performance. Journal of Personality andSocial Psychology, 17, 223–243.

Kaisa A., & Nurmi, J. (2005). The role of parenting styles in children’s problem behavior.Child Development, 76, 1144–1159.

Kochanska, G. & Aksan, N. (2006). Children’s conscience and self-regulation. Journal of Per-sonality, 74, 1587–1618.

Kochanska, G., Gross, J. N., Lin, M. H., & Nichols, K. E. (2002). Guilt in young children:Development, determinants, and relations with a broader system of standards. ChildDevelopment, 73, 461–483.

Lansford, J. E., Criss, M. M., Dodge, K. A., Shaw, D. S., Pettit, G. S., & Bates, J. E. (2009).Trajectories of physical discipline: Early childhood antecedents and developmentaloutcomes. Child Development, 80, 1385–1402.

Larzelere, R. E. (2000). Child outcomes of nonabusive and customary physical punishmentby parents: An updated literature review. Clinical Child and Family Psychology Review, 3,199–221.

Lepper, M. R. (1973). Dissonance, self-perception, and honesty in children. Journal of Personalityand Social Psychology, 25, 65–74.

Lepper, M. (1983). Social control processes, attributions of motivation, and the inter-nalization of social values. In E. T. Higgins, D. N. Ruble, &W. W. Hartup (Eds.), Social cog-nition and social development: A sociocultural perspective (pp. 294–330). Cambridge, England:Cambridge University Press.

Copyright © 2011 John Wiley & Sons, Ltd. Inf. Child Dev. 21: 310–322 (2012)DOI: 10.1002/icd

322 K. J. Rotenberg et al.

Lima, J., Caughy, M., Nettles, S. M., & O’Campo, P. J. (2010). Effects of cumulative risk onbehavioral and psychological well-being in first grade: Moderation by neighborhoodcontext. Social Science & Medicine, 71, 1447–1454.

Meltzer, H., Vostanis, P., Goodman, R., & Ford, T. (2007). Children’s perceptions ofneighbourhood trustworthiness and safety and their mental health. Journal of ChildPsychology and Psychiatry, 48, 1208–1213.

Oberwittler, D. (2007). The effects of neighbourhood poverty on adolescent problembehaviours: A multi-level analysis differentiated by gender and ethnicity. Housing Studies,22, 781–803.

Rotenberg, K. J. (1994). Loneliness and interpersonal trust. Journal of Social and ClinicalPsychology, 13, 152–173.

Rotenberg, K. J. (1995). The socialization of trust: Parents’ and their children’s interpersonaltrust. International Journal of Behavioural Development, 18, 713–726.

Rotenberg, K. J. (2010). The conceptualization of interpersonal trust: A basis, domain, andtarget framework. In Ken J. Rotenberg (Ed). Interpersonal trust during childhood andadolescence (pp 2–27), New York: Cambridge University Press.

Rotenberg, K. J., McDougall, P., Boulton, M. J., Vaillancourt, T., Fox, C. & Hymel, S. (2004).Cross-sectional and longitudinal relations among relational trustworthiness, socialrelationships, and psychological adjustment during childhood and adolescence in theUK and Canada. Journal of Experimental Child Psychology, 88, 46–67.

Rotenberg, K. J., Michalik, N., Eisenberg, N., & Betts, L. R. (2008). The relations amongyoung children’s peer-reported trustworthiness, inhibitory control, and preschool adjustment.Early Childhood Research Quarterly, 23, 288–298.

Shelton, K. K., Frick, P. J., & Wootton, J. (1996). Assessment of parenting practices in familiesof elementary school-age children. Journal of Clinical Child Psychology, 25, 317–329.

Smetana, J. G. (2006). Social-cognitive domain theory: Consistencies and variations inchildren’s moral and social judgments. In Melanie Killen (Ed) and Judith G Smetana(Ed.). Handbook of moral development (pp. 119–153). Mahwah, NJ, US: Lawrence ErlbaumAssociates.

Stemmler, M., Lösel, F., Beelmann, A., Jaursch, S., & Zenkert, B. (2005). Child problembehavior in kindergarten and in primary school: A comparison between predictionconfigural frequency analysis and multiple regression. Psychology Science, 47, 467–478.

Stilwell B. M., Galvin, M., Kopta, S. M., & Norton J. A. Jr. (1994). Moral–emotional respon-siveness: A two-factor domain of conscience functioning. Journal of the American Academyof Child And Adolescent Psychiatry, 33, 130–139

Stouthamer-Loeber, M. (1986). Lying as a problem behavior in children: A review. ClinicalPsychology Review, 6, 267–289.

Talwar, V., & Lee, K. (2008). Socio-cognitive correlates of children’s lying behaviour:Conceptual understanding of lying, executive functioning, and false beliefs.ChildDevelopment,79, 866–881.

Talwar, V., & Lee, K. (in press). A punitive environment fosters children’s dishonesty: Anatural experiment. Child Development.

Talwar, V., Gordon, H., & Lee, K. (2007). Lying in the elementary school: Verbal deceptionand its relation to second-order belief understanding. Developmental Psychology, 43,804–810.

Thompson, R. A., Meyer, S., & McGinley, H. (2006). Understanding values in relationships:The development of conscience. In M. Killen & J. G., Smetana (Eds.). Handbook of moraldevelopment (pp. 267–297). Mahwah, NJ, US: Lawrence Erlbaum Associates.

Tremblay, R. E., Loeber, R., Gagnon, C., Charlebois, P., Larivee, S. & LeBlanc, M. (1991).Disruptive boys with stable and unstable high fighting behavior patterns during juniorelementary school. Journal of Abnormal Child Psychology, 19, 285–300.

Vieno, A., Nation, M., Perkins, D. D., Pastore, M., & Santinello, M. (2010). Social capital,safety concerns, parenting, and early adolescents’ antisocial behaviour. Journal of CommunityPsychology, 38, 314–328.

Xu, F., Bao, X., Fu, X., Talwar, V., & Lee, K. (2010). Lying and truth-telling in children: Fromconcept to action. Child Development, 81, 581–596.

Copyright © 2011 John Wiley & Sons, Ltd. Inf. Child Dev. 21: 310–322 (2012)DOI: 10.1002/icd