15
British Journal of Educational Psychology (2014), 84, 253–267 © 2013 The British Psychological Society www.wileyonlinelibrary.com A person-centred analysis of teacherchild relationships in early childhood Kathleen Hughes, Amanda Bullock and Robert J. Coplan* Department of Psychology, Carleton University, Ottawa, Ontario, Canada Background. Previously, the nature of teacherchild relationships (TCRs) has been explored through dimensions of close, conflicted, and dependent. However, this variable-centred approach is limited, as many relationships can be characterized by more than one characteristic or trait. A person-centred approach would allow for a greater understanding of the types of relationships that are formed and improved interpretation of the students’ socio-emotional outcomes associated with that relationship. Aims. The primary goal of this study was to examine the socio-emotional functioning of young children who formed distinct types of TCRs. Sample. Participants were n = 202 kindergarten children (98 girls, 104 boys, M age = 64.12 months, SD = 4.86). Method. Multi-source assessment was employed with data collected from parent ratings, teacher reports, child interviews, and naturalistic observations in the classroom. Using quartile cut-offs, we identified children who formed conflicted, dependent, and ‘combined’ (conflicted and dependent) TCRs. Results. Results indicated distinct patterns of socio-emotional functioning for each TCR group. For example, whereas children in conflicted TCRs evidenced greater externalizing difficulties, children with dependent TCRs had greater internalizing difficulties. Children who evidenced high levels of both conflicted and dependent TCRs displayed the most pervasive socio-emotional difficulties. Conclusions. Children who form TCRs characterized by high levels of both conflict and dependency displayed the most pervasive adjustment difficulties. Further investigation is needed to improve our understanding of this group and to assess the plausibility of early intervention strategies. Teachers play a central part of children’s cognitive, academic, and socio-emotional development (Hamre & Pianta, 2001; Lerner, Lerner, & Zabski, 1985; O’Connor, Dearing, & Collins, 2011). For instance, teachers may serve as parent surrogates in their daily interaction with children. Teachers are frequently placed in caregiving roles and are responsible for listening to children’s concerns and easing upset feelings. The teacherchild relationship (TCR) is a focal relationship that may influence children’s socio- emotional development and behavioural functioning (Birch & Ladd, 1998; Roorda, Koomen, Spilt, & Oort, 2011; Rudasill & Rimm-Kaufmann, 2009). The nature of TCRs has been previously explored through dimensions of close, conflicted, and dependent (e.g., Pianta, Steinberg, & Rollins, 1995; Sabol & Pianta, 2012). The goal of the present study was *Correspondence should be addressed to Robert J. Coplan, Department of Psychology, Carleton University, 1125 Colonel By Drive, Ottawa, Ontario K1S 5B6, Canada (email: [email protected]). DOI:10.1111/bjep.12029 253

A person-centred analysis of teacher-child relationships in early childhood

Embed Size (px)

Citation preview

Page 1: A person-centred analysis of teacher-child relationships in early childhood

British Journal of Educational Psychology (2014), 84, 253–267

© 2013 The British Psychological Society

www.wileyonlinelibrary.com

A person-centred analysis of teacher–childrelationships in early childhood

Kathleen Hughes, Amanda Bullock and Robert J. Coplan*Department of Psychology, Carleton University, Ottawa, Ontario, Canada

Background. Previously, the nature of teacher–child relationships (TCRs) has been

explored through dimensions of close, conflicted, and dependent. However, this

variable-centred approach is limited, as many relationships can be characterized by more

than one characteristic or trait. A person-centred approach would allow for a greater

understanding of the types of relationships that are formed and improved interpretation

of the students’ socio-emotional outcomes associated with that relationship.

Aims. The primary goal of this study was to examine the socio-emotional functioning of

young children who formed distinct types of TCRs.

Sample. Participants were n = 202 kindergarten children (98 girls, 104 boys,

Mage = 64.12 months, SD = 4.86).

Method. Multi-source assessment was employed with data collected from parent

ratings, teacher reports, child interviews, and naturalistic observations in the classroom.

Using quartile cut-offs, we identified children who formed conflicted, dependent, and

‘combined’ (conflicted and dependent) TCRs.

Results. Results indicated distinct patterns of socio-emotional functioning for each TCR

group. For example, whereas children in conflicted TCRs evidenced greater externalizing

difficulties, children with dependent TCRs had greater internalizing difficulties. Children

who evidenced high levels of both conflicted and dependent TCRs displayed the most

pervasive socio-emotional difficulties.

Conclusions. Childrenwho formTCRs characterized by high levels of both conflict and

dependency displayed the most pervasive adjustment difficulties. Further investigation is

needed to improve our understanding of this group and to assess the plausibility of early

intervention strategies.

Teachers play a central part of children’s cognitive, academic, and socio-emotional

development (Hamre & Pianta, 2001; Lerner, Lerner, & Zabski, 1985; O’Connor, Dearing,

& Collins, 2011). For instance, teachers may serve as parent surrogates in their daily

interaction with children. Teachers are frequently placed in caregiving roles and areresponsible for listening to children’s concerns and easing upset feelings. The teacher–child relationship (TCR) is a focal relationship that may influence children’s socio-

emotional development and behavioural functioning (Birch & Ladd, 1998; Roorda,

Koomen, Spilt, & Oort, 2011; Rudasill & Rimm-Kaufmann, 2009). The nature of TCRs has

been previously explored through dimensions of close, conflicted, and dependent (e.g.,

Pianta, Steinberg,&Rollins, 1995; Sabol&Pianta, 2012). The goal of the present studywas

*Correspondence should be addressed to Robert J. Coplan, Department of Psychology, Carleton University, 1125 Colonel ByDrive, Ottawa, Ontario K1S 5B6, Canada (email: [email protected]).

DOI:10.1111/bjep.12029

253

Page 2: A person-centred analysis of teacher-child relationships in early childhood

to employ a person-centred approach to create distinct TCR groups and to examine the

socio-emotional outcomes associated with these groups.

Profiles of teacher–child relationships

Some TCRs are characterized by closeness and affection (O’Connor et al., 2011; Rudasill

&Rimm-Kaufmann, 2009). A closeTCRentails a high degree ofwarmth and support (Birch

& Ladd, 1998; Hamre & Pianta, 2001; O’Connor et al., 2011; Pianta, 1999). Children who

form close relationships with their teachers may be more likely to seek appropriate

assistance from teachers, respond to teachers’ requests, and be cooperative in the

classroom. Close TCRs have been associated with children’s positive socio-emotional

development. For example, close TCRs are associated with higher levels of socialcompetence (Pianta & Stuhlman, 2004; Sabol & Pianta, 2012) and lower levels of

aggressive and disruptive tendencies, lower social withdrawal, and greater prosocial

behaviour (Birch & Ladd, 1997; Ladd, Birch, & Buhs, 1999).

Conflicted TCRs may be primarily characterized by hostile or argumentative

interactions and a lack of rapport between the teacher and the child (Birch & Ladd,

1998; Hamre, Pianta, Downer, &Mashburn, 2007). Childrenmay challenge their teachers’

opinions and act uncooperative or defiant (Birch & Ladd, 1997). Highly conflicted TCRs

are associatedwith negative socio-emotional outcomes and behavioural problems such asaggression, hyperactivity, and classroom disruptions (Birch & Ladd, 1997; Hamre et al.,

2007; Pianta & Stuhlman, 2004; Silver, Measelle, Armstrong, & Essex, 2005). These ‘acting

out’ behavioural problems are considered to represent ‘externalizing’ difficulties

(Henricsson & Rydell, 2006; Ladd & Burgess, 1999). Conflict between teachers and

students in kindergarten has been linked to lower levels of prosocial behaviour and higher

levels of aggressive behaviour in the short term (Birch & Ladd, 1998) and aggression and

lower social competence throughout elementary school (Birch & Ladd, 1997; Pianta &

Stuhlman, 2004).Finally, dependent TCRs are typically characterized by a child’s over-reliance on

their teacher. Overly dependent children tend to behave in a possessive manner

towards their teacher (Birch & Ladd, 1997, 1998; Hamre & Pianta, 2001; Howes,

Phillipsen, & Peisner-Feinberg, 2000; Sabol & Pianta, 2012) and may rely on teachers to

resolve their socio-emotional or academic problems. In addition, children who form a

dependent relationship with their teacher may prefer interacting with their teachers as

compared to peers (Birch & Ladd, 1997). Children who are overly reliant on their

teachers may not understand the nuances of peer relationships and may be sociallywithdrawn and display lower social competence, loneliness, and anxiety (Howes et al.,

2000; Ladd & Burgess, 1999; Murray & Murray, 2004). Their withdrawing and ‘acting

inward’ problems are conceptualized as ‘internalizing’ difficulties (Henricsson & Rydell,

2006).

Recent TCR literature has moved away from examining dependent TCRs (O’Connor

et al., 2011; Roorda et al., 2011; Rudasill & Rimm-Kaufmann, 2009) through the

omission of the dependent subscale on the Student–Teacher Relationship Scale (STRS –Pianta, 1992). Instead, negativity in TCRs has been primarily measured through theconflicted subscale. However, the conflicted subscale assesses only negativity that is

commonly associated with externalizing problems. Recent studies such as that by

O’Connor et al. (2011) have found a lack of linear relations between conflicted TCRs

and internalizing difficulties. In recent years, peer relationships researchers have

argued for the importance of examining social withdrawal and internalizing difficulties

254 Kathleen Hughes et al.

Page 3: A person-centred analysis of teacher-child relationships in early childhood

in peer relations (Rubin & Coplan, 2004). Similarly, we argue that internalizing

difficulties are associated with TCRs, and the dependent subscale of the STRS should be

reconsidered.

Dimensions versus types

Previous research on TCRs has typically examined the dimensions of TCRs (closeness,

conflict, dependency) as independent variables (Rudasill & Rimm-Kaufmann, 2009). This

variable-centred approach has been used to examine the links between TCR character-

istics and child outcome variables (aggression, peer rejection, anxiety – Ladd & Burgess,

1999).

An alternative approach is to identify and describe distinct types of TCRs. Giventhat each relationship is likely characterized by multiple dimensions, considering

specific combinations of TCR characteristics may add to our understanding of how

different relationships are distinctly related to children’s socio-emotional functioning.

Such a person-centred approach allows for the grouping of participants based on

salient characteristics in their TCR and the examination of the outcomes related to

each ‘type’.

Some children may form a TCR that is characterized by a single salient characteristic.

For example, a conflicted TCR could be defined as high in teacher–child conflict, butrelatively average or low in dependency and closeness. Previous studies have used the

person-centred approach to examine single salient factors to compare children in a

high-conflicted group to children in a low-conflicted group (Silver et al., 2005). Other

studies have combined the close and conflicted subscales into a single factor and

compared trajectories of TCRs over time (O’Connor et al., 2011).

Alternatively, specific combinations of salient factors may also characterize a TCR. For

example, some children may form relationships that evidence high levels of both conflict

and dependency. Combinations of TCR groups were previously examined by Pianta(1994). To the best of our knowledge, no previous study has employed a person-centred

approach to examine TCR groups based on conflict and dependency.

The present study

The goal of this study was to examine the socio-emotional functioning of children who

formed different types of TCRs. Using a person-centred approach, we identified four

groups of children, whose TCRs were characterized as follows: (1) conflicted (highconflict, low-average dependency); (2) dependent (high dependency, low-average

conflict); (3) combined (high conflict, high dependency); and (4) a comparison group

(low-average conflict, low-average dependency).

Drawing upon the extant research (Ladd et al., 1999; O’Connor et al., 2011; Silver

et al., 2005), it was hypothesized that children with conflicted TCRs would be rated by

parents and teachers as havingmore externalizing problems, be less prosocial, and display

less social play behaviours as compared to children in a comparison group. It was

hypothesized that children with dependent TCRs would be rated as having moreinternalizing problems, be less prosocial, display less social play behaviours, displaymore

anxious play behaviours, and to report lower perceived competence and greater

loneliness as compared to children in a comparison group.

Children who are comorbid for aggression and social withdrawal have been

reported to have TCRs that are high in both conflict and dependency (Ladd &

Teacher–child relationships 255

Page 4: A person-centred analysis of teacher-child relationships in early childhood

Burgess, 1999). It was speculated that children who form relationships with teachers

which are combined for conflict and dependency may experience the most

substantive and extensive socio-emotional difficulties across all indices of socio-emo-

tional functioning.Finally, child gender was included as an additional predictor variable in all analyses.

Overall, teachers tend to report closer and less conflicted TCRs with girls than boys

(Furrer & Skinner, 2003; Hamre et al., 2007; Hughes, Cavell, & Willson, 2001; Roorda

et al., 2011).

Method

Participants

The participants in this study were 202 kindergarten children (98 girls, 104 boys;

Mage = 64.12 months, SD = 4.86) who attended 15 public schools located in a mid-sized

city in south-eastern Ontario, Canada. After obtaining ethics approval from both the

university Institutional Review Board and the local school board, principals and teachers

were solicited to participate. Child participants were recruited by sending letters of

informed consent home to parents from school. The sample was primarily Caucasian(70%) with other ethnicities also present (e.g., 17% Asian, 5% Black, 8% other).

Approximately 17% of mothers and 19% of fathers cited high school as their highest level

of education, 69% of mothers and 61% of fathers had a college/university degree, and 11%

of mothers and 16% of fathers had a graduate level degree.

Procedure

Data were collected over the course of a kindergarten school year. In January/February, teachers rated student–teacher relationships, and behavioural observations

of children’s play behaviours were collected. In May/June, children’s socio-emotional

functioning was assessed by parents and teachers, and child interviews.

Measures

Teacher ratings

Kindergarten teachers completed the STRS (Pianta, 1992). The STRS is a 28-itemmeasure

in which teachers rated their perceptions of their relationships with each student. The

STRS uses a 5-point Likert-type scale and contains three subscales including closeness

(11 items, e.g., ‘I share an affectionate, warm relationship with this child’), conflict(12 items, e.g., ‘This child and I always seem to be struggling with each another’), and

dependency (five items, e.g., ‘This child asks for my help when he/she does not need

help’). In the current sample, all three subscales displayed acceptable internal consistency

(conflict a = .91; closeness a = .84; dependency a = .74). The STRS has demonstrated

strong psychometric properties and evidenced predictive, concurrent, and construct

validity (Birch& Ladd, 1997; Hamre& Pianta, 2001; Howes et al., 2000; Murray&Murray,

2004; Pianta, 2001; Rudasill & Rimm-Kaufmann, 2009).

Teachers also completed the Child Behaviour Scale (Ladd & Profilet, 1996), a 35-itemmeasure of preschool children’s adjustment with peers with excellent psychometric

properties and strong validity (Ladd, 2006). Of particular interest were the subscales of

aggressive with peers (seven items, a = .91, e.g., ‘Kicks, bites or hits other children’),

256 Kathleen Hughes et al.

Page 5: A person-centred analysis of teacher-child relationships in early childhood

prosocial with peers (seven items, a = .90, e.g., ‘Kind towards peers’), and anxious with

peers (four items, a = .84, e.g., ‘Worries about many things’).

Behavioural observations

Observations of children’s behaviours were conducted by six research assistants.

Observers received extensive training and met regularly during data collection to discuss

any issues arising and to reduce rater drift. Children were observed and coded during free

play using an adapted versionof thePlayObservation Scale (Rubin, 2001). Each childwas

observed in 10-s intervals on at least three separate days for a total of 15 min (90 coding

intervals). For each coding interval, the child’s predominant free-play behaviour was

recorded (Coplan, Findlay, & Nelson,2004). For the present study, we examined anxiousplay (i.e., summed instances of on-lookingbehaviours andovert displays of anxious affect)

and social play (summed instances of group play and peer conversation; Coplan, Arbeau,

& Armer, 2008). Raw scores for each code were proportionalized for each child.

Inter-rater reliability based on 540 codes of data (90 min)was strong as the Cohen’s kappa

between observers ranged from .80 to .86.

Child interviews

Children were interviewed individually by research assistants to complete the

Loneliness and Social Dissatisfaction Questionnaire for Young Children (Cassidy

& Asher, 1992). Children responded with ‘yes’, ‘no’, or ‘sometimes’ to 16 items

related to loneliness and social dissatisfaction (a = .76, e.g., ‘Do you have kids to play

with at school?’). This measure has previously demonstrated excellent psychometric

properties, consistent factor structure across samples, high internal consistency, and

good test–retest reliability (Terrell-Deutsch, 1999). For the present sample, items were

averaged to create an aggregate measure with higher scores indicating greaterloneliness.

Children also completed thePictorial Scale for YoungChildren (Harter &Pike, 1984),

which is a widely used assessment of children’s perceptions of their competencies and

social relationships (Byrne, 1996). Of interest was the subscale of perceived competence

(12 items, a = .70), which assesses perceptions of child competencies (i.e., physical,

cognitive) in age-appropriate activities (e.g., puzzles, counting, running). For each item,

theparticipant is shown twopictures:One childwho is performingwell at a given activity,

and one child who is not performingwell. Participants are asked to choose which picturethey aremore like and then to point at either a large circle (‘I am a lot like this picture’) or a

small circle (‘I am a little like this picture’). This scale has been reported to have strong

psychometric properties (Harter & Pike, 1984) and has demonstrated good validity with

young children (Coplan et al., 2004).

Parent ratings

Mothers completed the Strengths and Difficulties Questionnaire (Goodman, 2001),which is a 3-point Likert scale consisting of 25 items of child behaviours. Of interest were

the subscales of children’s emotional symptoms, consisting primarily of items pertaining

to anxiety (five items, a = .74, e.g., ‘Often seemsworried’); conduct problems, consisting

primarily of items pertaining to aggression (five items, a = .70, e.g., ‘Often fights with

other children’); and prosocial (five items, a = .69, e.g., ‘Shares readily with other

Teacher–child relationships 257

Page 6: A person-centred analysis of teacher-child relationships in early childhood

children’). This scale is well validated as a brief screening tool for children’s behaviour

problems (Stone, Otten, Engels, Vermulst & Janssens, 2010).

Results

Preliminary analyses

Results from Little’s Missing Completely At Random test statistic indicated that the data

were not missing completely at random, v2(111) = 144.87, p < .05. Missing data were

under 9% for most variables with the exception of parent ratings, which had 22%missing

data. This was due to a higher proportion of parents not completing questionnaires.1

Accordingly, we conducted separate analyses for outcome variables as a function of

source of assessment (child-reported, observed, teacher-rated, parent-rated). No influen-

tial outliers (univariate or multivariate) were found. Descriptive statistics for all study

variables are displayed in Table 1.

The distribution of several variables violated the assumption of normality. Although

square root and inverse transformations were successful in normalizing the distributions,

this had no impact on themain analyses. Therefore, themain analyses are presented using

untransformed data.Correlations between all study variables are displayed in Table 2. Overall, the

pattern of association among variables was predominantly consistent with theoretical

expectations. For example, conflicted TCR was positively associated with

parent-rated conduct problems and negatively associated with perceived competence.

Also, prosocial behaviour was negatively correlated with aggressive and anxious

behaviour.

Age of childwas not significantly correlatedwith any study variables. Parent education

was significantly associated with teacher-rated aggression. Accordingly, results arepresented with parental education and without age included as a covariate in subsequent

MANCOVAs. Significant gender differences in TCRs were only noted for closeness, with

girls (M = 48.05, SD = 5.55) tending to form close relationships with teachers as

compared to boys (M = 45.24, SD = 6.72, t(192) = �3.17, p < .01).

Teachers’ ratings of conflicted and dependent TCRs were used to create four groups

based on quartile cut-offs. The conflicted group (n = 21; males = 15, female = 6)

included children in the top 25% on the conflicted scale and the bottom 75% on the

dependent scale. The dependent group (n = 25; males = 9, females = 16) was com-prised of children in the top 25% on the dependent scale and the bottom 75% on the

conflicted scale. The ‘combined’ group (n = 28; males = 16, females = 12) consisted of

children in the top 25% on both conflicted and dependent scales. Finally, the comparison

group (n = 120; male = 60, female = 60) included children in the bottom 75% on both

conflicted and dependent scales.2 Results from a chi-square analysis indicated that the

gender distribution of these groups did not differ significantly from expected values, v2(3,N = 194) = 6.21, p = .10.

1Children with missing parent-rated data had higher scores of externalizing behaviour (t = 11.38, p < .01) and TCRs conflict(t = 9.05, p < .01), and lower scores of TCR closeness (t = 5.44, p < .05) and social play (t = 4.38, p < .05) as comparedto other children.2Closeness was originally included in the grouping criteria, but the close group of participants (scoring in top 25% on close, bottom75%on conflict and dependent) were not distinguishable from the comparison group of participants scoring in the bottom75%onall scales. Very few participants (n = 7) were found to be in the top 25% of closeness in addition to the top 25% of dependencyand/or conflict. In addition, high dependency and conflict were associated with similar outcomes regardless of scores on closeness.Therefore, closeness was removed from the model.

258 Kathleen Hughes et al.

Page 7: A person-centred analysis of teacher-child relationships in early childhood

Comparison of teacher–child relationship groups

The goal of the main analyses was to explore differences in socio-emotional functioning

among different TCR groups. Accordingly, a series of 4 (TCR groups) by 2 (child gender)

MANCOVAs were conducted (with parental education serving as a covariate) to explore

differences in indices of socio-emotional functioning. Of note, no significantgender 9 TCR group interaction effects were found.

Observed play behaviours

The first MANCOVA examined the differences between TCR group and gender regarding

children’s observedplay behaviours (anxiousplay, social play,n = 178).Nomain effect of

gender was found, F(2, 171) = 0.98, p = .38, partial g2 = .01. However, results (Wilks’s

lambda) indicated a significantmain effect of TCRgroup,F(6, 342) = 3.20, p < .01, partialg2 = .05.

Follow-up univariate analyses revealed significant TCR group effects for social play, F

(3, 172) = 3.68, p < .05, partial g2 = .06, and anxious play F(3, 172) = 4.16, p < .01,

partial g2 = .07. Relevant means and SDs are presented in Table 3. Post-hoc analyses

(LSD) revealed that children in the TCR comparison group displayedmore social play than

children in the conflicted and combined group and less anxious play than children in the

conflicted and dependent groups. No other significant differences were found between

the other groups.

Child-reported adjustment

A second MANCOVA focused on child-rated measures of socio-emotional functioning

(loneliness, perceived competence, n = 184). There was no significant main effect of

Table 1. Descriptive statistics for all study variables

N Mean SD Range

Age (months) 200 64.12 4.86 48.00–76.00Parent education 201 3.41 0.86 1.50–5.00Teacher–child relationships

Close 194 46.60 6.32 24.00–55.00Conflicted 194 16.90 7.61 12.00–54.00Dependent 194 8.87 3.50 5.00–23.00

Observations

Anxious play 183 0.15 0.13 0.00–0.79Social play 183 0.56 0.21 0.00–1.00

Child self-reports

Loneliness 189 1.36 0.31 1.00–2.75Perceived competence 189 3.31 0.41 1.33–4.00

Teacher ratings

Anxiety 190 1.25 0.43 1.00–3.00Aggression 192 1.31 0.41 0.68–3.00Prosocial 192 2.25 0.48 1.00–3.00

Parent ratings

Emotional symptoms 157 1.55 1.70 0.00–7.00Conduct problems 157 1.30 1.31 0.00–6.00Prosocial 157 8.06 1.62 4.00–10.00

Teacher–child relationships 259

Page 8: A person-centred analysis of teacher-child relationships in early childhood

Table

2.Associationsbetw

eenstudyvariables

23

45

67

89

10

11

12

13

14

15

1.C

hild

age

�0.03

0.09

�0.06

�0.02

�0.13

0.09

�0.06

0.04

�0.02

0.08

0.01

�0.12

�0.03

�0.04

2.Parenteducation

–�0

.16*

�0.14

0.19**

0.02

�0.02

�0.13

0.03

�0.11

�0.20**

0.11

�0.06

�0.07

�0.01

3.C

onflictedST

R–

0.52***

�0.37***

0.03

�0.18*

0.14

�0.20**

0.17*

0.55***

�0.32***

0.05

0.27**

�0.25**

4.D

ependentST

R–

�0.09

0.10

�0.13

0.16*

�0.07

0.33***

0.35***

�0.10

0.17*

0.10

�0.18*

5.C

lose

STR

–�0

.08

0.18*

�0.13

0.10

�0.05

�0.21**

0.28***

�0.19*

0.01

0.14

6.A

nxiousplay(O

)–

�0.40***

0.09

�0.06

0.19*

�0.16*

�0.05

0.03

0.06

�0.05

7.Socialplay(O

)–

�0.16*

�0.02

�0.26**

�0.05

0.09

0.01

0.03

�0.05

8.Loneliness

(C)

–�0

.24**

0.11

0.13

�0.11

0.20*

0.04

�0.10

9.C

ompetence

(C)

–�0

.07

�0.13

0.09

�0.13

�0.02

0.06

10.A

nxious(T)

–0.22**

�0.14

0.10

�0.02

�0.01

11.A

ggressive(T)

–�0

.37***

�0.01

0.20*

�0.20**

12.Prosocial(T)

–0.10

�0.19*

0.16*

13.Emotional(P)

–0.14

�0.10

14.C

onduct

(P)

–0.37***

15.Prosocial(P)

Note.ST

R=student–teacherrelationships;O

=observations;C

=child

report;T

=teacherreport;P

=parent–report.

*p<.05;**p

<.01;***

p<.001.

260 Kathleen Hughes et al.

Page 9: A person-centred analysis of teacher-child relationships in early childhood

gender, F(2, 177) = 0.26, p = .78, partial g2 < .01. However, results indicated a main

effect of TCR group, F(6, 354) = 3.05, p < .01, partial g2 = .05. Significant univariate

effects for TCR group were found regarding loneliness, F(3, 175) = 3.13, p < .05, partial

g2 = .05, and perceived competence, F(3, 175) = 3.48, p < .05, partial g2 = .06. Results

indicated that children in the comparison group reported higher perceived competence

than children in the conflicted and combined groups. In addition, children in the

combined group reported being lonelier than children in the comparison and conflicted

groups (see Table 3).

Teacher-rated adjustment

The third MANCOVA examined the teacher-rated measures of child behaviour (anxious,

aggressive, prosocial, n = 188). Results indicated significant main effects of TCR group, F

(9, 438) = 8.45,p < .001, partialg2 = .12, and gender, F(3, 180) = 7.73,p < .001, partial

g2 = .11. Follow-up univariate analyses indicated significant TCR group effects for

teacher-rated anxiety, F(3, 182) = 5.35, p < .01, partial g2 = .08, aggression, F(3,182), = 21.20, p < .001, partial g2 = .26, and prosocial behaviour, F(3, 182) = 4.55,

p < .01, partial g2 = .07. Results indicated that children in the comparison group were

rated as less anxious than children in the dependent and combined groups. Children in the

combined group were rated as more aggressive than children in the comparison,

dependent, and conflicted groups. Children in the conflicted group were rated as more

aggressive than the comparison and dependent groups. Finally, children in the combined

group were rated as less prosocial than the comparison and dependent groups. Also, the

dependent group was rated higher in prosocial behaviour than children in the conflictedgroup (see Table 3).

Significant univariate gender effects were found for aggression, F(1, 182) = 15.50,

p < .001, partial g2 = .08, and prosocial behaviour, F(1, 182) = 12.15, p < .01, partial

Table 3. Estimated means (standard errors) of outcome variables among teacher–child relationship

groups

Teacher–Child relationship group

Comparison Conflicted Dependent Combined

Outcome variable

Observations

Social play 0.60a (0.20) 0.46b (0.22) 0.51ab (0.26) 0.49b (0.17)

Anxious play 0.13a (0.11) 0.19b (0.14) 0.23b (0.20) 0.14ab (0.12)

Child reports

Loneliness 1.32a (0.30) 1.31a (0.22) 1.37ab (0.29) 1.54b (0.41)

Competence 3.38a (0.36) 3.12b (0.37) 3.30ab (0.46) 3.17b (0.54)

Teacher reports

Anxiety 1.16a (0.33) 1.32ab (0.54) 1.45b (0.52) 1.44b (0.52)

Aggression 1.19a (0.26) 1.55b (0.40) 1.25a (0.36) 1.82c (0.57)

Prosocial 2.30ab (0.49) 2.06ac (0.39) 2.44b (0.42) 1.96c (0.37)

Parent reports

Emotional 1.43a (1.60) 1.08a (1.44) 1.64ab (1.62) 2.69b (2.09)

Conduct problems 1.20a (1.30) 1.38ab (1.26) 0.91a (1.19) 2.25b (1.24)

Prosocial 8.23a (1.52) 7.85ab (1.52) 8.32ab (1.70) 7.25b (1.91)

Note. Means with different subscripts in the same row differ significantly at the .05 level.

Teacher–child relationships 261

Page 10: A person-centred analysis of teacher-child relationships in early childhood

g2 = .06, but not anxiety, F(1, 182) = 1.96, p = .16, partial g2 = .01. Results indicated

that boys were rated as displaying more aggression and less prosocial behaviours as

compared to girls.

Parent-rated adjustment

The final MANCOVA focused on parent-rated measures of child behaviour (emotional

symptoms, conduct problems, prosocial, n = 153). Results indicated a significant main

effect of TCR group, F(9, 353) = 2.21, p < .05, partial g2 = .04, and gender, F(3,

145) = 4.00, p < .01, partial g2 = .08.

Follow-up univariate analyses indicated significant TCR group effects for parent-rated

conduct problems, F(3, 147) = 3.83, p < .05, partial g2 = .07, and emotional symptoms,F(3, 147) = 2.96, p < .05, partial g2 = .06. Results indicated that children in the

combined group were rated as displaying more conduct problems than children in the

comparison or dependent groups. Notably, conflicted children did not differ from any

other group on this measure. Also, children in the combined group were rated as

displayingmore emotional symptoms than the comparison and conflicted groups, but not

the dependent group. No TCR group effects were found regarding prosocial behaviour, F

(3, 147) = 1.85, p = .14, partial g2 = .04 (see Table 3).

Significant univariate gender effects were found regarding parent-rated prosocialbehaviour, F(1, 147) = 7.78, p < .01, partial g2 = .05. Results indicated that girls were

rated as more prosocial than boys. No significant gender effects were found regarding

parent-rated conduct problems, F(1, 147) = 0.73, p = .40, partial g2 < .01, or emotional

problems, F(1, 147) = 0.45, p = .33, partial g2 < .01.

Discussion

The goal of this studywas to examine socio-emotional outcomes associatedwith different

types of TCRs in early childhood. Using a person-centred approach, we compared groups

of kindergarten children who formed conflicted and dependent TCRs. In addition, we

identified of a group of combined children, whose TCRs were characterized as both

dependent and conflicted. To our knowledge, this was the first study to examine such a

group. These different types of TCRs were compared in terms of teacher-rated and

parent-rated behaviour problems, child self-reports of loneliness and perceived compe-tence, as well as observed play behaviours. Our results indicated unique socio-emotional

profiles for each type of TCR group.

Conflicted teacher–child relationships

Children who formed primarily conflicted TCRs were rated by teachers as being more

aggressive as compared to children in the dependent and comparison TCR groups.

Interestingly, conflicted children did not differ from the other TCR groups in terms ofparent-rated conduct problems. These findings provide partial support for previous

research which found significant associations between conflicted TCRs and external-

izing behaviour as rated by mothers and teachers (O’Connor et al., 2011; Pianta &

Stuhlman, 2004; Silver et al., 2005). This lack of association with parental ratings may

be due to missing parent data, as missingness of the data was linked with teacher-rated

externalizing behaviour. Alternatively, this may also be due to a disparity between

262 Kathleen Hughes et al.

Page 11: A person-centred analysis of teacher-child relationships in early childhood

parent and teaching ratings of child behaviours (Hinshaw, Han, Erhardt, & Huber,

1992).

Children in the conflicted group were rated by teachers as less prosocial as compared

to children in the dependent group, but did not differ from children in the comparisongroup. This provides mixed support for previous research, which found a negative

association between conflicted TCRs and prosocial behaviour (e.g., Hamre et al., 2007).

Children in the conflicted group reported lower ratings of perceived competence

compared with children in the comparison group. In addition, children in the conflicted

group displayed more anxious play and less social play compared with children in the

comparison group. These findings may indicate that children in the conflicted group lack

the necessary social skills to form positive and prosocial relationships (Zimmer-Gembeck,

Geiger, & Crick, 2005).Despite these peer problems, conflicted children were not found to be at heightened

risk for internalizing problems. Children in the conflicted group did not differ from the

comparison group in terms of parent-rated emotional symptoms, teacher-rated anxiety, or

child-reported loneliness. Thus, conflicted children demonstrated some specificity in

terms of their socio-emotional difficulties (Howes et al., 2000; Pianta & Stuhlman, 2004).

Dependent teacher–child relationshipsChildren who formed dependent TCRs were rated by teachers as more anxious than the

comparison group. In addition, children in the dependent groupwere observed to display

more anxious play than children in the comparison group. These findings may indicate

that children who form dependent TCRs may be more likely to experience internalizing

problems (Birch&Ladd, 1997, 1998;Howes et al., 2000;Murray&Murray, 2004; Pianta&

Nimetz, 1991; Pianta et al., 1995). In contrast, dependent children did not differ from the

comparison group in terms of parent-rated conduct problems or teacher-rated aggression,

suggesting that these childrenwere not perceived as aggressive or hostile by their parentsor teachers.

Notwithstanding, dependent children also did not differ significantly from other

groups in terms of parent-rated emotional symptoms, self-reported loneliness or

perceived competence, or observed social play. These findings were not consistent with

previous research, which found that dependent TCRs were associated with child

self-reported loneliness and other internalizing difficulties (Birch & Ladd, 1997, 1998;

Murray & Murray, 2004; Pianta & Nimetz, 1991; Pianta et al., 1995). Our findings may be

due to the methodology with which current data were collected. For instance, childrenwho form dependent TCRs may not display the same degree of internalizing problems at

home, when they are in a comfortable and familiar environment (Spooner, Evans, &

Santos, 2005).

Combined teacher–child relationships

Children in the combined group had both highly conflicted and highly dependent TCRs.

The present study was the first to focus on this particular group of children. Our findingssuggest this group of children may very well merit our additional attention. As compared

to children in other TCR groups, children in the combined group evidenced the most

pervasive deficits in socio-emotional functioning. For example, combined children were

rated by both parents and teachers as having more internalizing (emotional symptoms,

anxiety) and externalizing (conduct, aggression) problems than comparison children, and

Teacher–child relationships 263

Page 12: A person-centred analysis of teacher-child relationships in early childhood

as being less prosocial. Indeed, teachers rated children in the combined group as higher in

aggression than any other group. In addition, children in the combined group reported

being lonelier and as having lower perceived competence than comparison children.

Finally, combined children were also observed to engage in less social play than theircomparison counterparts.

Taken together, these findings do not ‘bode well’ for children who form TCRs

characterized by high levels of both conflict and dependency. To begin with, negative

TCRs in early childhood predict developmental risk factors such as academic

difficulties and behavioural problems (Furrer & Skinner, 2003; Hamre & Pianta,

2001). In addition, a lack of social play in early childhood is a risk factor for both

internalizing and externalizing problems (Coplan & Arbeau, 2009). Early internalizing

problems may predict later mental health problems (Goodwin, Fergusson, & Horwood,2004), and early externalizing problems may predict later delinquency, substance

abuse, or antisocial behaviours (Zimmer-Gembeck et al., 2005). Although most young

children have relatively high self-esteem, those with comparatively negative self-

perceptions are at risk of negative developmental outcomes (Coplan et al., 2004).

Finally, comorbidity of behaviour problems is often associated with greater maladjust-

ment than a single diagnosis (Ladd & Burgess, 1999; Wilens et al., 2002). Accordingly,

our findings suggest that children in the combined TCR group are at a heightened risk

for a number of developmental risk factors. Our results support cause to alerteducators and researchers to pay attention to this group. Future research is needed to

improve our understanding of this group and to assess the plausibility of early

intervention strategies.

Limitations and future directions

This study examined TCRs using a person-centred approach. Through grouping children

based on the salient qualities of their relationships with their teachers, we compared thesocio-emotional outcomes associated with four TCR groups. Our findings indicated that

childrenwith dependent and conflicted TCRsmay be at particular risk of internalizing and

externalizing problems, respectively. However, children who form relationships with

their teachers which are characterized with both conflict and dependency may be at the

greatest risk of socio-emotional difficulties in early childhood.

Notwithstanding, several caveats in this research should be considered. Our

correlational study does not permit the assessment of directional effects or causation.

Future research should examine the stability and change in TCR groups over time, as wellas longer-term implications. Moreover, a wider range of outcome variables should be

explored, including assessments of academic functioning.

As well, having teachers rate both TCR quality and child behaviour problems

introduces issues with regard to shared-method variance. Teachers’ perceptions of their

students may have influenced both their ratings of the TCR and their ratings of student

behaviour towards peers. Although our findings were strongest when considering

teacher-rated outcomes, some significant differences between TCR groups (particularly

regarding combined children) were evident among parent-rated, child-reported, andobserved child behaviours. Notably, findings concerning parent ratings may have been

underestimated due to the large number of missing data.

Further, the current study only examined groups based upon conflicted and

dependent TCRs. Closeness was excluded from the grouping criteria as a ‘close’ group,

and the comparisons were found to be indistinguishable. In addition, high dependency

264 Kathleen Hughes et al.

Page 13: A person-centred analysis of teacher-child relationships in early childhood

and high conflictwere associatedwith similar outcomes regardless of scores on closeness.

Thus, closeness was excluded from the current study model. This lack of effect of

closeness may be due to the small sample size. Therefore, future studies are needed to

investigate close TCRs with a larger sample.Notwithstanding, the current study provided new insight into the connections

between TCRs and socio-emotional outcomes in early childhood. The person-centred

approach allowed for the exploration of TCR groups and the identification of children

with highly conflicted and highly dependent TCRs. These findings may help in the

identification of children who may warrant attention and might merit from targeted early

intervention in school.

Acknowledgements

This research was supported by a Social Science and Humanities Research Council of

Canada grant to author Coplan. The authors wish to thank Madelena Arnone, Belinda

Boekhoven, Leanna Closson, Brooke Fletcher, Sherri Frohlick, Allison Graham, and

Deborah McEwen for their help in the collection and coding of data.

References

Birch, S. H., & Ladd, G. W. (1997). The teacher-child relationship and children’s early school

adjustment. Journal of School Psychology, 35, 61–79. doi:10.1016/S0022-4405(96)00029-5Birch, S. H., & Ladd, G. W. (1998). Children’s interpersonal behaviours and the teacher-child

relationship. Developmental Psychology, 34, 934–946. doi:10.1037/0012-1649.34.5.934Byrne, B. M. (1996). Measures of self-concept for young children. In B. M. Byrne (Ed.), Measuring

self-concept across the life span: Issues and instrumentation (pp. 69–84). Washington, DC:

American Psychological Association.

Cassidy, J., & Asher, S. R. (1992). Loneliness and peer relations in young children. Child

Development, 63, 350–365. doi:10.2307/1131484Coplan, R. J., & Arbeau, K. (2009). Peer interactions and play in early childhood. In K. H. Rubin,

W. Bukowski & B. Laursen (Eds.), Handbook of peer interactions, relationships, and groups

(pp. 143–161). New York: Guilford.

Coplan, R. J., Arbeau, K. A., & Armer, M. (2008). Don’t fret, be supportive!Maternal characteristics

linking child shyness to psychosocial and school adjustment in kindergarten. Journal of

Abnormal Child Psychology, 36 , 359–371. doi:10.1007/s10802-007-9183-7Coplan, R. J., Findlay, L. C., & Nelson, L. J. (2004). Characteristics of preschoolers with lower

perceived competence. Journal of Abnormal Child Psychology, 32 , 399–408. doi:10.1023/B:JACP.0000030293.81429.49

Furrer, C., & Skinner, E. (2003). Sense of relatedness as a factor in children’s academic engagement

and performance. Journal of Educational Psychology, 95, 148–162. doi:10.1037/0022-0663.95.1.148

Goodman, R. (2001). Psychometric properties of the Strengths and Difficulties Questionnaire.

Journal of the American Academy of Child and Adolescent Psychiatry, 40, 1337–1345.doi:10.1097/00004583-200111000-00015

Goodwin, R. D., Fergusson, D. M., & Horwood, J. (2004). Early anxious/withdrawn behaviours

predict later internalising disorders. Journal of Child Psychology and Psychiatry, 45, 874–883.doi:10.1111/j.1469-7610.2004.00279.x

Hamre, B. K., & Pianta, R. C. (2001). Early teacher-child relationships and the trajectory of children’s

school outcomes through eighth grade. Child Development, 72, 625–638. doi:10.1111/

1467-8624.00301

Teacher–child relationships 265

Page 14: A person-centred analysis of teacher-child relationships in early childhood

Hamre, B. K., Pianta, R. C., Downer, J. T., & Mashburn, A. J. (2007). Teachers’ perceptions of

conflict with young students: Looking beyond problem behaviours. Social Development, 17

(1), 115–136. doi:10.1111/j.1467-9507.2007.00418.xHarter, S., & Pike, R. (1984). The pictorial scale of perceived competence and social acceptance for

young children. Child Development, 55, 1969–1982. doi:10.2307/1129772Henricsson, L., & Rydell, A. (2006). Children with behaviour problems: The influence of social

competence and social relations on problem stability, school achievement, and peer

acceptance across the first six years of school. Infant and Child Development, 15, 347–366.doi:10.1002/icd.448

Hinshaw, S. P., Han, S. S., Erhardt, D., & Huber, A. (1992). Internalizing and externalizing behaviour

problems in preschool children: Correspondence among parent and teacher ratings and

behaviour observations. Journal of Clinical Child Psychology, 21, 143–150. doi:10.1207/s15374424jccp2102_6

Howes, C., Phillipsen, L. C., & Peisner-Feinberg, E. (2000). Consistency of perceived teacher–childrelationships between preschool and kindergarten. Journal of School Psychology, 38,

113–132. doi:10.1016/S0022-4405(99)00044-8Hughes, J. N., Cavell, T. A., &Willson, V. (2001). Further support for the developmental significance

of the student-teacher relationship. Journal of School Psychology, 39, 289–301. doi:10.1016/S0022-4405(01)00074-7

Ladd, G. W. (2006). Peer rejection, aggressive or withdrawn behaviour, and psychological

maladjustment fromages 5 to 12: Anexamination of fourpredictivemodels.ChildDevelopment,

77 , 822–846. doi:10.1111/j.1467-8624.2006.00905.xLadd, G. W., Birch, S. H., & Buhs, E. S. (1999). Children’s social and scholastic lives in kindergarten:

Related spheres of influence? Child Development, 70, 1373–1400. doi:10.1111/1467-8624.00101

Ladd,G.W.,&Burgess, K. B. (1999). Charting the relationship trajectories of aggressive,withdrawn,

and aggressive/withdrawn children during early grade school. Child Development, 70 , 910–929. doi:10.1111/1467-8624.00066

Ladd, G. W., & Profilet, S. M. (1996). The Child Behaviour Scale: A teacher-report measure of

young children’s aggressive, withdrawn, and prosocial behaviours.Developmental Psychology,

32, 1008–1024. doi:10.1037/0012-1649.32.6.1008Lerner, J. V., Lerner, R. M., & Zabski, S. (1985). Temperament and elementary school children’s

academic performance: A test of ‘goodness-of-fit’ model. Journal of Child Psychology, 26, 125–136. doi:10.1111/j.1469-7610.1985.tb01633.x

Murray, C., & Murray, K. M. (2004). Child level correlates of student-teacher relationships:

An examination of demographic characteristics, academic orientations, and behavioural

orientations. Psychology in the Schools, 41, 751–762. doi:10.1002/pits.20015O’Connor, E. E., Dearing, E., & Collins, B. A. (2011). Teacher-child relationships and behaviour

problem trajectories in elementary school. American Education Research Journal, 48 (1),

120–162. doi:10.3102/0002831210365008Pianta, R. C. (1992). The Student Teacher Relationship Scale. Charlottesville, VA: University of

Virginia.

Pianta, R. C. (1994). Patterns of relationships between children and kindergarten teachers.

Journal of School Psychology, 32, 15–31. doi:10.1016/0022-4405(94)90026-4Pianta, R. C. (1999). Enhancing relationships between children and teachers. Washington, DC:

American Psychological Association.

Pianta, R. C. (2001). Student-Teacher Relationship Scale. Lutz, FL: Psychological Assessment

Resources.

Pianta, R. C., & Nimetz, S. (1991). Relationships between children and teachers: Associations

with classroom and home behaviour. Journal of Applied Developmental Psychology, 12,

379–393. doi:10.1016/0193-3973(91)90007-Q

266 Kathleen Hughes et al.

Page 15: A person-centred analysis of teacher-child relationships in early childhood

Pianta, R. C., Steinberg, M. S., & Rollins, L. B. (1995). The first two years of school: Teacher-child

relationships and deflections in children’s classroom adjustment. Development and

Psychopathology, 7, 295–312. doi:10.1017/S0954579400006519Pianta, R. C., & Stuhlman, M. W. (2004). Teacher-child relationships and children’s success in the

first years of school. School Psychology Review, 33 , 444–458.Roorda, D. L., Koomen, H. M. Y., Spilt, J. L., & Oort, F. J. (2011). The influence of affective

teacher-student relationships on student’s school engagement and achievement: Ameta-analytic

approach. Review of Education Research, 81 , 493–529. doi:10.3102/0034654311421793Rubin, K. H. (2001). The Play Observation Scale (POS). College Park, GA: University of Maryland.

Rubin, K. H., & Coplan, R. J. (2004). Paying attention to and not neglecting social withdrawal

and social isolation. Merrill-Palmer Quarterly, 50, 506–534. doi:10.1353/mpq.2004.0036

Rudasill, K. M., & Rimm-Kaufmann, S. E. (2009). Teacher-child relationship quality: The roles of

child temperament and teacher-child interactions. Early Childhood Research Quarterly, 24,

107–120. doi:10.1016/j.ecresq.2008.12.003Sabol, T. J., & Pianta, R. C. (2012). Recent trends in research on teacher-child relationships.

Attachment & Human Development, 14 , 213–231. doi:10.1080/14616734.2012.672262Silver, R. B., Measelle, J. R., Armstrong, J. M., & Essex, M. J. (2005). Trajectories of classroom

externalising behaviour: Contributions of child characteristics, family characteristics, and the

teacher-child relationship during the school transition. Journal of School Psychology, 43, 39–60. doi:10.1016/j.jsp.2004.11.003

Spooner, A. L., Evans,M. A., & Santos, R. (2005). Hidden shyness in children: Discrepancies between

self-perceptions and the perceptions of parents and teachers. Merrill-Palmer Quarterly, 51,

437–466. doi:10.1353/mpq.2005.0028

Stone, L. L., Otten, R., Engels, R. C., Vermulst, A. A., & Janssens, J. M. (2010). Psychometric

properties of the parent and teacher versions of the strengths and difficulties questionnaire for 4-

to 12-year-olds. Clinical Child Family Psychology Review, 13, 254–274. doi:10.1353/mpq.

2005.0028

Terrell-Deutsch, B. (1999). The conceptualization and measurement of childhood loneliness.

In K. J. Rotenberg & S. Hymel (Eds.), Loneliness in social isolation. Cambridge, MA: MIT Press.

Wilens, T. E., Biederman, J., Brown, S., Tanguay, S., Monuteaux, M. C., Blake, C., & Spencer, T. J.

(2002). Psychiatric comorbidity and functioning in clinically referred preschool children and

school-age youths with ADHD. Journal of the American Academy of Child & Adolescent

Psychiatry, 41 , 262–268. doi:10.1097/00004583-200203000-00005Zimmer-Gembeck, M. J., Geiger, T. C., & Crick, N. R. (2005). Relational and physical aggression,

prosocial behaviour, and peer relations: Gender moderation and bidirectional associations.

The Journal of Early Adolescence, 25, 421–452. doi:10.1177/0272431605279841

Received 20 July 2012; revised version received 15 August 2013

Teacher–child relationships 267