Teaching and Teacher Education 53 (2016) 30–40

Contents lists available at ScienceDirect

Teaching and Teacher Education

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

Social dynamics in the classroom: Teacher support and conflict and

the peer ecology
Marloes M.H.G. Hendrickx a,∗, M.Tim Mainhard a, Henrike J. Boor-Klip b,
Antonius H.M. Cillessen b, Mieke Brekelmans a
a
Utrecht University, Department of Education, The Netherlands
b
Radboud University Nijmegen, Behavioural Science Institute, The Netherlands

h i g h l i g h t s

• Teacher support is positively related to overall peer liking and prosocial behavior.
• Teacher conflict is positively related to peer disliking and aggressive behaviors.
• The more differentially teachers behave, the more hierarchical the peer ecology is.
• The more support teachers provide, the less hierarchical the peer ecology is.

a r t i c l e i n f o a b s t r a c t

Article history: By showing support and conflict, teachers may function as a model for students regarding how to inter-
Received 1 May 2015 act and how to evaluate each other, thereby shaping the classroom peer ecology. Associations of general
Received in revised form 18 August 2015
and student-specific levels and differential provision of teacher support and conflict with the classroom
Accepted 13 October 2015
peer ecology were investigated. Multivariate multiple regression analyses were performed with a sample
Available online 28 October 2015
of 58 Dutch fifth-grade classrooms (1454 students). In particular student perceptions of teacher support
Keywords: and conflict, rather than teacher perceptions or observations, explained peer liking and disliking, the de-
Teacher support gree of social hierarchy, and how prosocial versus aggressive the peer ecology was.
Teacher conflict © 2015 The Authors. Published by Elsevier Ltd.
Differential teacher behavior This is an open access article under the CC BY-NC-ND license
Peer ecology (http://creativecommons.org/licenses/by-nc-nd/4.0/).

1. Introduction dents’ development (Gest & Rodkin, 2011; Rodkin & Hodges, 2003).
Nonetheless, there has been little research on associations between
The classroom peer ecology, or the social environment of class- teacher behavior and peer relations. Because of this relative lack of
room peers in interaction with each other, is one of the most im- research, Farmer, McAuliffe Lines, and Hamm (2011) referred to the
portant proximal environments for students’ social (Ahn, Rodkin, teacher’s influence on peer relations as “the invisible hand” of the
& Garandeau, 2010; Farmer & Xie, 2007; Hodges, Boivin, Vitaro, teacher. The few studies conducted so far (e.g., Hughes, Cavell, &
& Bukowski, 1999; Roland & Galloway, 2002) and academic de- Willson, 2001; McAuliffe, Hubbard, & Romano, 2009) have mainly
velopment (Kindermann, 2007; Roseth, Johnson, & Johnson, 2008; examined how teacher–student interactions and relationships are
Wentzel & Caldwell, 1997). Teachers, as professionals in a posi- associated with the position of specific students within the class-
tion very close to the peer group, may have a role in shaping the room group, and have not investigated how teachers may influ-
nature of their classroom’s peer ecology. Affecting the peer ecol- ence the classroom peer ecology as such. Some characteristics of
ogy deliberately may even be a strategy for teachers to foster stu- the peer ecology, such as the social structure or status hierarchy,
only exist at this classroom-level and cannot be grasped when fo-
cusing on student-level outcomes. In only one study, Gest and Rod-
kin (2011) examined associations between general teacher prac-
∗
Corresponding author. Department of Education, Utrecht University, P.O. Box tices and the peer ecology of the entire classroom group. In the
80140, 3508 TC Utrecht, The Netherlands.
present study, we aim to further reveal the teacher’s invisible hand
E-mail addresses: m.m.h.g.hendrickx@uu.nl (M.M.H.G. Hendrickx),
m.t.mainhard@uu.nl (M.Tim Mainhard), h.klip@psych.ru.nl (H.J. Boor-Klip), by examining how teacher support and conflict are related to the
a.cillessen@psych.ru.nl (A.H.M. Cillessen), m.brekelmans@uu.nl (M. Brekelmans). nature of the classroom peer ecology.

http://dx.doi.org/10.1016/j.tate.2015.10.004
S0742-051X(15)30013-5/© 2015 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
 M.M.H.G. Hendrickx et al. / Teaching and Teacher Education 53 (2016) 30–40 31

To study the complexity of peer relations within the social sys- we focus on both positive (prosocial) and negative (aggressive) be-
tem of a class, we employed a social network approach. Social net- haviors. Two of the most basic prosocial behaviors are cooperating
work analysis is used not just to count the number of ties between and helping (Rubin et al., 2006). Aggression has often been sub-
peers in a class, but also to examine in more detail patterns or divided into overt aggression (hitting, calling each other names)
structures of relationships (e.g., hierarchy) among individuals in a and relational aggression (gossiping, excluding others; Crick &
group (Borgatti, Everett, & Johnson, 2013). Grotpeter, 1995). As the prevalence or commonness of such be-
havior describes what is currently normal behavior in a group, we
1.1. Classroom peer ecology use the term descriptive norm (see Chang, 2004; Lapinski & Ri-
mal, 2005). Next to current commonness of behaviors, classroom
The concept of peer ecology is rooted in ecological systems the- descriptive norms are associated with future prevalence of behav-
ory (Bronfenbrenner, 1979; Bronfenbrenner & Morris, 2006), which iors, as social behaviors tend to be contagious (Dishion & Piehler,
describes how an individual is nested within social settings, like 2009); in classrooms where aggression is the norm, students tend
families or classrooms. Interactions the child has within a setting, to conform to this norm and become more aggressive themselves
called proximal processes, are considered to be “primary mecha- (Espelage, Holt, & Henkel, 2003; Thomas, Bierman, & the Conduct
nisms producing human development” (Bronfenbrenner & Morris, Problems Prevention Research Group, 2006). Furthermore, descrip-
2006, p. 795). Thus, by interacting with each other, children in- tive norms predict how strongly behaviors are associated with ac-
fluence and socialize each other. A set of individuals in interac- ceptance or rejection (Chang, 2004).
tion is referred to as a social microsystem (Neal & Neal, 2013),
within which Bronfenbrenner and Morris (2006) further distin- 1.2. Teacher support and conflict
guished between patterns of interpersonal relations, social roles,
and activities. This distinction was used in research on peer re- Given the importance of the classroom peer ecology as a so-
lationships (Gest & Rodkin, 2011) to describe the classroom peer cial context for students’ development, it is necessary for teach-
ecology as encompassing (a) the richness of interpersonal ties, (b) ers to understand how they may, unwillingly or deliberately, affect
social structure or status hierarchy, and (c) patterns of social be- these ecologies. Gest and Rodkin (2011) suggested that the teacher,
haviors exhibited by classroom peers (see also Rubin, Bukowski, & who has a position close to the peer group, is the one profes-
Parker, 2006). In the present study, we examine these three aspects sional who has the opportunity to oversee and affect the class-
of classroom peer ecologies. room peer ecology. Gest and Rodkin developed a model of how
teacher practices affect students’ individual development, partly
1.1.1. Richness of interpersonal ties
through affecting the classroom peer ecology. They described how
The richness of interpersonal ties indicates how many positive
the peer ecology is impacted both by everyday teacher–student
and how few negative relationships are present among students
interactions and by “network-related teaching”, that is, conscious
in a classroom. Following a long history of research on peer rela-
teaching strategies directly aimed at affecting peer relationships.
tionships (e.g., Coie, Dodge, & Coppotelli, 1982; Rubin et al., 2006),
In the present study the focus is on everyday teacher–student in-
we focus on liking and disliking. In classrooms where many stu-
teraction, and more specifically on the amount of support and con-
dents like each other well, students are more likely to feel secure
flict in teacher–student relationships and interactions. Teacher sup-
and accepted, which in turn positively affects academic adjustment
port, or warmth, fosters individual students’ social (e.g., Luckner &
(Roseth et al., 2008; Wentzel & Caldwell, 1997). Larger numbers of
Pianta, 2011; Verschueren, Doumen, & Buyse, 2012) and academic
positive ties in a classroom also imply less negative behavior like
adjustment (e.g., Cornelius-White, 2007; Den Brok, Brekelmans, &
bullying (Roland & Galloway, 2002).
Wubbels, 2004), whereas teacher conflict amplifies externalizing
1.1.2. Status hierarchy behavior (e.g., Doumen et al., 2008; Runions, 2014) and is nega-
The status hierarchy refers to the degree to which social sta- tively related to academic achievement (e.g., Ladd, Birch, & Buhs,
tus in the classroom peer ecology is structured in an egalitarian 1999; Mantzicopoulos, 2005).
versus hierarchical manner. In the case of an egalitarian distribu- In daily classroom practice, teachers interact both with individ-
tion, each student has a similar status, whereas in a hierarchical ual students and with the classroom group as a whole. Wubbels
peer group a small number of students has a relatively high peer et al. (2015) argued that teachers differ in the extent to which
status and is in that sense more socially dominant (Brown, 2011). they establish warm, supportive relationships at these two levels.
At the student level, likeability and popularity reflect two aspects A teacher who shows much support to individual students may
of peer status (Cillessen, 2011). Whereas likeability is a combina- not be able to establish supportive interactions during whole-class
tion of how well a student is liked by every other individual in the teaching. Another teacher may convey much warmth or support
classroom, popularity refers to a student’s visibility, dominance, or when teaching the class as a whole, but may keep individual inter-
prestige and thus more directly reflects a position in the peer ecol- action formal and less supportive. Corresponding to these concep-
ogy (Cillessen, 2011). When likeability or popularity is distributed tually different levels, studies have either investigated teacher sup-
highly hierarchically in the class, only some students are liked by port and conflict with a specific student as the object (e.g., Hughes
the majority of their peers or are considered to be highly popular. et al., 2001; Verschueren et al., 2012) or as more general – in the
Schäfer, Kron, Brodbeck, Wolke, and Schulz (2005) found that with sense of not student-specific – teacher or classroom characteristics
a more pronounced status hierarchy, there was more negative be- (e.g., Luckner & Pianta, 2011; Mainhard, Brekelmans, & Wubbels,
havior – in their study tapped by bullying and victimization – than 2011). A study by Den Brok, Brekelmans, and Wubbels (2006) il-
in classrooms where social status was distributed more equally. lustrates the relevance of distinguishing student-specific from gen-
Furthermore, a study by Cappella, Kim, Neal, and Jackson (2013) eral teacher support and conflict; Den Brok et al. found qualitative
showed that students in classes with a more egalitarian structure differences in teacher support, depending on whether the class as
were more behaviorally engaged than students in classes with less a whole or individual students were the focus in otherwise similar
network equity. questionnaire items.
The present study adopts this distinction between general and
1.1.3. Social behavior student-specific teacher support and conflict. Although student-
A third aspect of the peer ecology is the social behaviors that specific teacher support and conflict are first and foremost ori-
characterize daily interactions among peers. In the current study ented at the individual student, these can be informative about
32 M.M.H.G. Hendrickx et al. / Teaching and Teacher Education 53 (2016) 30–40

a teacher’s classroom practices in interactions with their students 1.3.1. Richness of interpersonal ties
in (at least) two different ways. First, classroom-average measures Through general support versus conflict, teachers may model
of student-specific support and conflict indicate how a teacher positive versus negative interpersonal relationships. As a result,
generally behaves with students in dyadic interaction (e.g., Buyse, students in classes with relatively higher levels of general teacher
Verschueren, Verachtert, & Van Damme, 2009; Hughes, Zhang, & support are expected to form more liking and less disliking rela-
Hill, 2006). Hughes et al. (2006) referred to this aggregate as the tionships. In line with this notion, Gest and Rodkin (2011), in a U.S.
classroom norm of support and conflict, which resonates with the sample of first, third and fifth-grade classes, found that teachers
classroom descriptive norms of student behaviors as discussed in who showed high levels of general emotional support had class-
Section 1.1.3. Second, it may also be worthwhile to examine the rooms with more reciprocated friendships.
extent of teachers’ differential provision of support and conflict. Classroom norms of student-specific support are also expected
Research on teacher differential behavior has its origins in stud- to result in a peer ecology that is richer in positive ties, since
ies on the teacher-expectancy effect (Rosenthal & Jacobson, 1968), students are more likely to be viewed by their peers in a posi-
which states that some teachers treat students differently based tive light. Similarly, in a classroom in which a teacher has many
on the level of achievement they expect of the student (see Babad, conflicted relationships with individual students, the peer ecol-
2009). One aspect of teacher–student interaction in which teach- ogy is likely to be characterized by more negative ties as stu-
ers treat students differentially is the amount of support or affect dents learn to approach many students negatively. In a Belgian
they show, which is typically higher for high-expectancy students study, Buyse et al. (2009) indeed found that first-grade classroom
(Babad, 2009) and also for students whom teachers feel closer to norms of student-specific teacher support were positively related
(Newberry & Davis, 2008). So, in the present study, student-specific to third-grade peer liking, whereas first-grade classroom norms of
teacher support and conflict are used as the building blocks of student-specific teacher conflict were negatively associated with
classroom norms of support and conflict and of teacher differen- third-grade peer liking. In the same vein, Hughes et al. (2006)
tial behavior. showed that the classroom norm of supportive relationships was
positively related to the average amount of peer liking in the class-
room group in first and second grade in the U.S.

1.3. Teacher support and conflict and the peer ecology 1.3.2. Status hierarchy
We expect social status hierarchy to be mainly related to teach-
Two mechanisms describe how teachers’ general and student- ers’ differential behavior; when teacher support or conflict is
specific support and conflict may relate to the peer ecology, being highly focused on a small group of students, the teacher informs
modeling and social referencing. We first elaborate on these mech- the students on the differential value of these peers. Hughes, Im,
anisms and then relate them to the three aspects of peer ecolo- and Wehrly (2014) have studied the impact of teacher differential
gies as introduced above. First, teachers’ general social practices in provision of support on peer experiences in third and fourth grade
class can be a model for peer interactions and peer relationships. in the U.S. They reasoned that when the provision of support is
In this view, teacher support or conflict set the tone for, or model, more egalitarian, more students have the opportunity to be per-
peer interactions in the classroom and communicate information ceived positively by their peers, leading to less hierarchy in the
about the types of interactions and relationships that students are peer-ecology. Hughes et al. found that when supportive relation-
expected to establish with each other (Farmer et al., 2011; Gest ships were concentrated on just a few students, also more status
& Rodkin, 2011). When teachers generally show support and have hierarchy in peer relations occurred. This was true however, specif-
positive interactions, the modeling perspective assumes that stu- ically for students’ academic reputation as an outcome rather than
dents are likely to emulate this behavior, that is, to show warmth for peer liking.
to each other and engage in positive interactions with peers as Next to differential provision of individual teacher support and
well. Likewise, teachers who generally show much conflict and conflict, general teacher support may also be related to the hier-
negative affect may stimulate conflicted contact among students archy in ties. Chang (2003) showed that in Chinese middle-school
as well (Farmer et al., 2011; Mikami, Griggs, Reuland, & Gregory, classrooms where teachers rated themselves as showing relatively
2012). more warmth, peers disliked withdrawn and especially aggressive
Whereas the modeling perspective emphasizes how students students less than in classrooms where teachers deemed them-
take in their teachers’ general support and conflict as implicit selves as being less warm. A study by Cappella and Neal (2012),
lessons for how to behave themselves, the social referencing per- with second to fourth-grade students, also showed that general
spective focuses on how students implicitly learn how to evalu- teacher support can buffer negative relationships of victims of bul-
ate and approach a specific student, depending on the teacher’s lying. A generally supportive teacher may thus relieve negative
student-specific provision of support and conflict (Buyse et al., peer evaluations of neglected or rejected students, and thereby fos-
2009; Hughes & Chen, 2011). Hughes et al. (2001) were the first to ter a more egalitarian peer ecology.
reason that the teacher functions as a social referent in the class-
room, that is, that “classmates make inferences about children’s 1.3.3. Social behavior
attributes and likeability based, in part, on their observations of By modeling supportive interactions in general, teachers com-
teacher–student interactions” (p. 289). The social referencing prin- municate to their students the social value of prosocial interac-
ciple applies to both norms of teacher support and conflict and tion (Farmer et al., 2011). Luckner and Pianta (2011) have found
teacher differential behavior. That is, when a teacher shows sup- that general teacher support was positively related to students’
port to many individual students and thereby sheds a positive light prosocial behaviors in a sample of fifth-grade students in U.S. el-
on each of them, this may result in peers learning how to view ementary schools. Similarly, teacher conflict may function as a
specific students more positively, which in turn may lead to a more model for students’ antisocial, aggressive behaviors. Furthermore,
pleasant peer ecology. When teachers differentially treat students we expect that in a classroom with a higher classroom norm of
and focus their positive (or negative) comments on only a few stu- student-specific support, prosocial behavior receives more atten-
dents, they inform the classroom group on their peers’ differential tion, whereas in classrooms with a higher norm of student-specific
value (Mikami, Lerner, & Lun, 2010), which may result in a more conflict, aggressive behaviors are addressed more. This may add to
hierarchical peer ecology. the students’ perception of the degree to which these behaviors
 M.M.H.G. Hendrickx et al. / Teaching and Teacher Education 53 (2016) 30–40 33

2. Method

2.1. Participants

Teachers and their students in 58 fifth grade classrooms in 40

Dutch primary schools participated in this study, which was part
of a larger research project on the classroom climate in fifth grade
of elementary school. 1 In the Netherlands, elementary school stu-
dents have the same teacher for every lesson (approximately 25 h
a week), or two teachers who each work part-time. When a class
had two teachers, the teacher who spent most hours in the class-
room participated in the study (60.3% of the participating teach-
ers spent at least 4 days a week with the class). Teachers were on
average 41.25 years old (range 24.51–62.47, SD = 11.91); teachers’
Fig. 1. Overview of the study constructs. General, classroom-based teacher support
and conflict are hypothesized to function as a model for the relationships and inter- mean experience was 15.17 years (range 1–39, SD1 = 10.99) and
actions students have in the classroom peer ecology. Student-specific teacher sup- 37 teachers were women (63.8%).
port and conflict, both the average level and the differential behavior of the teacher, Class size ranged from 18 to 34 students (M = 26.16, SD = 3.70).
are expected to function as a social referent, providing information about students Only students for whom informed parental consent was obtained
within the peer ecology.
participated (1496 out of 1518, participation rate 98.6%). On the
day of data collection, 42 students were absent, resulting in a sam-
ple size of 1454. Of these students, 47.4% were girls, and class-
rooms contained 30.4%–66.7% girls. Students’ mean age was 10.59
are normative, and as a result to their own exhibition of the be-
years (SD = .49). Based on the classification by Statistics Nether-
havior. In the study by Buyse et al. (2009), first-grade classroom
lands (2012b), 84.2% of the students were Dutch (both parents
normative conflict was indeed positively associated with aggressive
were born in The Netherlands) and classes had 0%–80% students
student behaviors.
who were not Dutch (at least one parent was not born in The
Netherlands). This distribution was representative for the areas in
1.4. The present study which the schools were located (Statistics Netherlands, 2012a).

In the present study associations between teacher support and

2.2. Measures of peer ecology
conflict and the classroom peer ecology are examined. By focusing
on these class-level constructs, we aimed to investigate the social
Peer ecology measures were based on peer nominations of like-
structures that define the setting within which students develop.
ability and popularity status and social behaviors. Participants an-
More specifically, we examined three aspects of teacher support
swered several questions about which of their classmates best fit-
and conflict, being (a) the level of general support and conflict, (b)
ted a certain description. With a minimum of one, they could nom-
classroom norms of student-specific teacher support and conflict,
inate as many classmates as they wanted, whether or not present
and (c) teachers’ differential provision of student-specific support
and whether or not consented. We excluded nominations of non-
and conflict. Also three aspects of peer ecologies were examined,
consented students from the dataset. To avoid sequence effects
being (a) richness of interpersonal ties, (b) hierarchy of these ties,
(Poulin & Dishion, 2008), the order of the classmates’ names from
and (c) classroom norms of social behaviors. An overview of these
which participants could choose was different for each participant.
constructs is provided in Fig. 1.
Both when a teacher models support or conflict in general
(Farmer et al., 2011; Gest & Rodkin, 2011) and when the teacher 2.2.1. Richness of interpersonal ties
is a social referent for the social evaluation of specific students Students completed nominations about peers whom they liked
by peers (Hughes et al., 2001), the teacher’s influence on the peer most (“Which classmates do you like most?”), and peers whom
ecology seems to depend on the students’ intake of teacher behav- they liked least (“Which classmates do you like least”). To indi-
ior. Therefore, we made sure to incorporate the students’ views on cate to what degree liking and disliking ties were present in the
all aspects of teacher support and conflict. For triangulation pur- classroom, density of both types of ties was calculated. That is, we
poses, a multiple informants design was used, including the stu- divided the total number of nominations within the classroom by
dents’ as well as the teacher’s or an external observer’s perspective the maximum possible nominations, being m∗(n−1), where m is
for each of the three aspects of teacher support and conflict. The the number of individuals who providing ratings and n is the num-
overarching research question was: How are teacher support and ber of consented students in the classroom (Wasserman & Faust,
conflict associated with the classroom peer ecology? 1994). The resulting density scores lie between 0 and 1 and rep-
Given what we discussed so far, we expected that teachers resent the degree to which liking and disliking ties are present in
would model peer interactions and relationships in general but the classroom. A liking density score close to 0 means that very
would also be a social referent for the evaluation of specific stu- few students indicated that they liked others, whereas a score of
dents. We expected that the more support and the less conflict 1 indicates that all students indicated that they liked all of their
teachers showed in general and to specific students, (a) the more peers.
classrooms would be characterized by positive rather than nega-
tive ties, and (b) the more prosocial behaviors would prevail over
2.2.2. Status hierarchy
aggressive behaviors. Furthermore, more differential provision of
The degree to which social status was hierarchically distributed
teacher support was expected to be related to more status hi-
in a classroom was based on the distribution of nominations for
erarchy in classrooms, since then the teacher would specifically
highlight differences between students. Also, we expected that the
more general support a teacher would show, the more egalitarian 1
One classroom participating in the larger study was excluded from these results,
the classroom distribution of ties would be. because of a diverting teaching approach with two teachers in the class at all times.
34 M.M.H.G. Hendrickx et al. / Teaching and Teacher Education 53 (2016) 30–40

the sociometric item measuring liking and an item measuring pop- To assess the reliability of the group averages, we computed the
ularity: “Which classmates are most popular?”. We calculated in- ICC2 (see Lüdtke, Robitzsch, Trautwein, & Kunter, 2009), which was
degree centralization for each of the status items to represent .87, indicating a highly reliable class-mean rating. For the teachers’
how hierarchically versus equally nominations were distributed. self-perception, the Cronbach’s alpha was .77.
In-degree refers to the number of nominations received by an in-
dividual, and centralization refers to the level of concentration of
these nominations on one or a few students. In-degree centraliza- 2.3.2. Classroom norms of student-specific teacher support and
tion is the difference in number of received nominations between conflict
the most nominated student and all the others. The formula pre- Sociometric items were used to tap peer teacher relationship
sented by Wasserman and Faust (1994) was used: In-degree cen- reputation (cf. Hughes et al., 2001), a measure for student-specific
tralization = sum[max(Pi)-Pi]/((m−1)∗(n−1)), where max(Pi) is the teacher–student relationships from a peer perspective. We used
largest number of nominations received by anyone in the class- two items measuring peer reputation of teacher support (“Which
room, Pi is the number of nominations received by an individual, m classmates are liked most by the teacher?” and “Which classmates
is the number of individuals providing ratings and n is the number are praised a lot by the teacher?”, r = .71, p < .001) and two
of consented students in the classroom. The centralization scores items to tap peer reputation of teacher conflict (“Which classmates
lie between 0 and 1, with a higher score representing a higher de- are liked least by the teacher?” and “Which classmates does the
gree of status hierarchy. The lowest score of 0 means that all class- teacher often get angry with?”, r = .92, p < .001). For each stu-
mates have equal status. The highest score of 1 indicates that only dent a score for peer reputation of teacher support as well as con-
one classmate has very high status, whereas all the others have flict was calculated as the average number of received nominations
very low status. for the two items. To obtain classroom norms of student-specific
teacher support and conflict as perceived by peers, we calculated
2.2.3. Social behavior density scores for each of the composite scores in the same man-
For every classroom, we calculated the descriptive norms of ner as the peer ecology density scores were calculated.
prosocial behavior and aggression as the density of peer nomina- Video observations were used to measure normative student-
tions for each of the behaviors. The prosocial items were “Which specific teacher support and conflict from an observer’s perspec-
classmates cooperate well?”, and “Which classmates help other tive. We coded all public dyadic teacher–student interactions for
children?” (r = .75, p < .001). Aggression comprised both overt the amount of support and conflict as expressed by the teacher.
aggression: “Which classmates call other children names?”, and Public dyadic teacher–student interactions were those interactions
“Which classmates hit or kick other children?”, and relational ag- noticeable for at least half the classroom and as expressed to or
gression: “Which classmates gossip about other children?” and about a single student or a small group, that is, those students
“Which classmates exclude other children?”. The Cronbach’s alpha whose name was called or a small group of collaborating stu-
for these four items was .86. For each student composite scores dents (maximum four students, e.g., “the group over there”, “the
were calculated for each type of behavior as the average number green group”). Each occurrence received a coded that represented
of received nominations on the relevant items. Next, as an indica- the amount of support or conflict, ranging from −2 (very low,
tor of the commonness of each type of behavior in a class (i.e., the i.e., conflict) to +2 (very high, i.e., support). The classroom norm
classroom norm) density scores were computed using the formula of student-specific teacher support was calculated as the mean
that was presented above. score of all instances of teacher–student interaction throughout the
video.
2.3. Measures of teacher support and conflict Table 1 shows an overview of the operationalization and ex-
amples for each score. The first author and two trained research
2.3.1. General teacher support and conflict assistants scored the videos. Inter-observer agreement was first
Both teachers and students completed the Questionnaire on checked for video segmentation; agreement that an event had oc-
Teacher Interaction for Primary Education (QTI-PE), a measure that curred ranged from 81% to 87% for the pairs of observers. Next, a
targets teacher support and conflict in general and combines the set of 1624 occurrences of teacher–student interaction (9% of the
two as opposite ends of a single dimension: communion. The QTI- total number of fragments) was coded by all three coders to test
PE is an adaptation of the QTI as developed for secondary ed- for inter-observer agreement. Weighted Cohen’s kappa ranged from
ucation (Wubbels, Brekelmans, Den Brok, & van Tartwijk, 2006). .69 to .76 for the pairs of observers, which can be considered sub-
For this younger population, the questionnaire contained short- stantial agreement (Landis & Koch, 1977).
ened items, and words that had proven to be difficult to read or
understand were replaced with synonyms. The questionnaire con-
sists of 16 items which each reflect a certain degree of communion. 2.3.3. Differential teacher behavior
The item “This teacher is friendly”, for example, reflects a high de- As an indication of differential teacher behavior as perceived
gree of communion (i.e., support), and items such as “This teacher by the students, we calculated in-degree centralization of the peer
yells” and “This teacher is impatient” reflect a low level of commu- reputation of teacher support and conflict nominations in the same
nion (i.e., conflict). A 5-point Likert-scale was used, ranging from 1 manner as for the status hierarchy scores. A higher score repre-
(almost never) to 5 (almost always). Students completed the ques- sented a higher degree of differential behavior. The lowest score of
tions about their teacher, and teachers indicated how they per- 0, for example for centralization of nominations for teacher sup-
ceived their own teaching in this class. For every class, items stated port, means that all classmates are perceived by their peers to re-
the name of the teacher involved. Following standard procedures ceive an equal amount of support. The highest score of 1 indicates
(Wubbels et al., 2006) each item was weighted for the degree of that only one classmate is perceived to receive teacher support.
communion and the sixteen weighted item scores were averaged, To tap differential teacher behavior from an observer’s per-
resulting in a single score for each teacher, ranging between minus spective, first per student the mean support score in all the ob-
1 and plus 1. For the student version, the Cronbach’s alpha was .81. served interactions the teacher had with that student was com-
Individual students’ reports of teacher support were averaged per puted. The standard deviation of the classroom mean of these in-
classroom. The intra-class correlation (ICC1), or the average agree- dividual scores represented the degree of teacher differential be-
ment between a pair of students within the same class, was .21. havior.
 M.M.H.G. Hendrickx et al. / Teaching and Teacher Education 53 (2016) 30–40 35

Table 1
Operationalization of observed student-specific teacher support.

Score Level of teacher support Indicators

−2 Very low (i.e., conflict) Angry or hostile

Sarcasm
“Stop that!”
“Don’t … !”
“You are being really annoying right now!”
−1 Low Voice is not louder than normal
“Please stop that”
“Could you sit normally?”
“I don’t like that”
Warning a child by calling their name
0 Intermediate/no information Organizational comments
“What is the answer to question 8?”
“Sally will work with Ellen”
“That is correct”
1 High Showing compassion
Smiling
“That’s nice of you”
“Thank you”
2 Very high (i.e., support) Words of affection, like sweetheart, darling, dear, my friend
Laughing and joking
“That is very nice of you!”
“Thank you very much!”

2.4. Procedure 3. Results

Data were collected in the fall semester of 2012/2013, at least 3.1. Peer ecology
one month after the start of the school year. Schools located in
the middle, south, and east of the Netherlands were recruited to Table 2 shows the descriptive statistics for the peer ecology
participate. After a school’s principal and the classroom teacher variables. Peer ecologies on average were more positive than neg-
agreed to participate, parents received information about the goals ative in nature; density for liking was higher than for dislik-
of the study and were asked for their consent regarding the partic- ing, t(57) = 8.78, p < .001, and prosocial behavior nominations
ipation of their child. All students for whom consent was granted were more common than nominations of aggression, t(57) = 18.30,
completed the questionnaires on netbook computers in their own p < .001. Popularity was to a higher degree centralized than liking,
classrooms. Students could not skip questions themselves. Yet, if t(57) = 20.47, p < .001.
they wanted to pass over a question or stop participating, they There was a positive association between liking and disliking
could inform the researcher and were allowed to do so. Students density, r = .52, p < .001. This correlation was less strong but still
sat separately and had partition screens flanking the netbooks to apparent when controlling for class size, r = .46, p < .001. Ap-
prevent distraction and to increase students’ privacy. A standard parently, in classrooms where students indicated liking for more
instruction was given concerning the content of the questions and peers, students also indicated disliking for more peers. Hierarchy
confidential data handling. Teachers also completed their question- in likeability and popularity were positively correlated, r = .47,
naires on a netbook. In addition, 2 h of video were recorded on the p < .001. Furthermore, liking hierarchy was positively related to
same day the questionnaires were completed. During the observa- liking density, r = .32, p = .013. So, in classrooms where students
tion teachers were free to follow their normal lesson plans. We indicated that they liked more peers, these nominations were more
did ask them not to schedule tests, because generally hardly any strongly focused on a small group of students. Finally, prosocial be-
interaction takes place during tests, and individual presentations, havior and aggression were uncorrelated when class size was con-
because interaction then typically revolves around the presenting trolled for, r = .15, p = .273.
student, resulting in an unrepresentative amount of differential be-
havior. After all data were collected, teachers received a summary
3.2. Teacher support and conflict
of the findings for their classrooms.
Table 3 shows the descriptive statistics for the teacher support
2.5. Analysis and conflict variables. Teachers perceived themselves as generally
less supportive than their students did, t(57) = −6.25, p < .001.
First, we checked the data for normality, multicollinearity and Both teachers and students saw more general teacher support than
(multivariate) outliers and found no violations. Data were then conflict, as both values were above 0, t(57) = 14.82, p < .001 and
analyzed using multivariate multiple regression analysis in MPlus t(57) = 28.48, p < .001 for teachers and students, respectively.
version 7.2 (Muthén & Muthén, 2014). Because of the limited sam- Consistently, student-specific teacher support was perceived by the
ple size, not all peer ecology measures could be tested within a students to be more common than teacher conflict, t(57) = 13.97,
single model. Instead, three separate models were tested, one for p < .001. However, the average observed support score was be-
each of the peer ecology measures. These models included all cor- low 0, t(56) = −2.22, p = .030, so according to the coders teachers
relations among teacher support and conflict measures. Finally, we expressed somewhat more conflict than support in their student-
controlled for class size, since density and centralization measures specific interactions. Finally, support was less centralized on a
are not completely independent of the number of students provid- small group of students than conflict, t(57) = 13.76, p < .001.
ing nominations (cf. Gest & Rodkin, 2011). Moderate positive correlations were found between the teach-
ers’ and students’ perception of the amount of general teacher sup-
36 M.M.H.G. Hendrickx et al. / Teaching and Teacher Education 53 (2016) 30–40

Table 2
Descriptive statistics for peer ecology measures.

1 2 3 4 5 M SD Min Max

Richness of ties
1 Liking density – .14 .04 .06 .29
2 Disliking density .52∗∗ – .10 .03 .04 .17
Status hierarchy
3 Liking hierarchy .32∗ .03 – .19 .05 .11 .34
4 Popularity hierarchy .11 −.03 .47∗∗ – .55 .15 .20 .84
Social behavior
5 Prosocial behavior .60∗∗ .36∗∗ .26∗ .09 – .27 .07 .13 .44
6 Aggressive behavior .24 .72∗∗ −.15 −.13 .34∗∗ .12 .04 .06 .26
∗ ∗∗
p < .05. p < .01.

Table 3
Descriptive statistics for teacher support and conflict measures.

1 2 3 4 5 6 7 M SD Min Max

General
1. Students’ perspective – .47 .13 .15 .69
2. Teacher’s perspective .46∗∗ – .34 .18 −.25 .72
Student-specific
3. Support norm .34∗∗ .19 – .29 .09 .11 .56
4. Conflict norm −.24 −.26∗ .11 – .12 .03 .06 .20
5. Observed support norm .35∗∗ .22 .32∗ −.08 – −.05 .17 −.47 .37
Differential behavior
6. Differential support .09 −.23 −.07 .23 .12 – .29 .08 .10 .50
7. Differential conflict .02 −.21 −.06 −.00 −.08 −.02 – .61 .15 .25 .92
8. Differentiality observed −.25 −.01 −.20 −.04 −.05 .04 .06 .31 .16 .05 .85
∗ ∗∗
p < .05. p < .01.

port were positively correlated, r = .46 p < .001 and between ob- conflict liking nominations tended to be distributed more equally,
servations and student ratings of student-specific teacher support, β = −.33, p = .028. Furthermore, as expected, when teachers
r = .31 p = .018. General and student-specific teacher support and according to students provided support more differentially, lik-
conflict were also correlated in the expected directions; the more ing nominations were more hierarchically distributed, β = .32,
supportive the students perceived the teacher to be in general, p = .015. Surprisingly, hierarchy in popularity was not associated
the more student-specific teacher support they reported, r = .34 with any of the teacher support measures. The model explained
p = .009, and the more student-specific supportive interactions 19.0% and 5.5% of the variance in hierarchy in liking and popular-
were observed, r = .35 p = .007. Furthermore, teacher-perceived ity, respectively.
general support was negatively related to classroom norms of peer-
perceived teacher conflict, r = −.26 p = .046. The measures of dif-
3.3.3. Social behaviors
ferential teacher behavior were not related to one another or to
As shown in the right part of Table 4, in classrooms with
the measures of general or normative teacher support.
higher norms of student-specific teacher support, significantly
more prosocial behavior was reported by students, β = .63,
3.3. Teacher support and conflict and the classroom peer ecology
p < .001. Surprisingly, also when teacher support was provided
more differentially (i.e., to a larger extent centralized on only a few
Finally, the multivariate multiple regression analyses were per-
students) more prosocial behavior was reported, β = .22, p = .027.
formed, one for each aspect of the peer ecology.
In classrooms where students reported higher norms of student-
specific teacher conflict, students reported more aggressive behav-
3.3.1. Richness of interpersonal ties
ior, β = .45, p < .001. The model explained 55.2% and 49.8% of the
The left part of Table 4 shows the results for the measures of
variance in prosocial behavior and aggression, respectively.
richness of interpersonal ties. As expected, in classrooms where
students nominated more peers to receive teacher support, stu-
dents also indicated liking for more peers, β = .66, p < .001. Also 4. Discussion
as expected, in classrooms where students indicated more student-
specific teacher conflict, students reported disliking for more of In the present study associations between teacher support and
their peers, β = .43, p = .001. Surprisingly, disliking density was conflict and the classroom peer ecology were examined. By focus-
also positively associated with classroom norms of student-specific ing at the peer ecology at the classroom level, we had the oppor-
support, β = .33, p = .005. The overall model explained 37.5%, and tunity to investigate how teacher support and conflict are related
45.0% of the variance in liking density and disliking density, re- to the social structure of a class within which student learning
spectively. and development occur (Bronfenbrenner & Morris, 2006). General
and student-specific teacher support and conflict were expected to
3.3.2. Status hierarchy function as a model (see Farmer et al., 2011) and a social referent
The middle section of Table 4 shows the results for the sta- for peer relationships (see Hughes et al., 2001), showing students
tus hierarchy variables. Classrooms where students perceived more how to behave in social interaction in general and how to evaluate
general teacher support had a more egalitarian liking structure, and approach specific peers. Next to overall levels of support and
β = −.32, p = .042. Contrary to this finding and to the expecta- conflict, we included differential behavior in our models, in order
tions, also in classes with higher norms of student-specific teacher to tap how equally teachers divided their attention, both positive
 M.M.H.G. Hendrickx et al. / Teaching and Teacher Education 53 (2016) 30–40 37

Table 4
Richness of Interpersonal Ties, Status Hierarchy, and Social Behavior as explained by Teacher Support and Conflict.

Richness of interpersonal ties Status hierarchy Social behavior

Liking density Disliking density Liking hierarchy Popularity hierarchy Prosocial behavior Aggression

B (SE) β B (SE) β B (SE) β B (SE) β B (SE) β B (SE) β

General
1. Students’ perspective −.06 (.05) −.18 .01 (.03) .03 −.13 (.07)∗∗ −.32 .01 (.20) .01 −.12 (.06) −.22 .07 (.04) .24
2. Teacher’s perspective −.01 (.03) −.02 −.03 (.02) −.14 .02 (.05) .08 −.09 (.14) −.10 .03 (.04) .08 −.04 (.03) −.17
Student-specific
3. Support norm .30 (.06)∗∗ .66 .11 (.04)∗∗ .33 .13 (.08) .24 .17 (.25) .11 .46 (.08)∗∗ .63 .07 (.05) .19
4. Conflict norm −.17 (.20) −.11 .45 (.14)∗∗ .43 −.61 (.28)∗∗ −.33 -.91 (.86) −.18 .01 (.27) .00 .58 (.16)∗∗ .45
5. Obs. support norm −.04 (.03) −.17 .00 (.02) .02 .02 (.04) .07 .04 (.13) .05 .01 (.04) .03 .01 (.02) .04
Differential behavior
6. Differential support .03 (.06) .06 .01 (.04) .03 .20 (.09)∗ .32 .00 (.26) .00 .18 (.08)∗∗ .22 .02 (.05) .05
7. Differential conflict .00 (.03) −.00 .00 (.02) −.02 −.00 (.04) −.00 −.14 (.13) −.14 −.02 (.04) −.05 .04 (.02) .15
8. Differentiality obs. .02 (.03) .08 −.02 (.02) −.08 −.06 (.04) −.18 .03 (.14) .03 −.01 (.03) −.03 −.01 (.03) −.05
R2 .37∗∗ .45∗∗ .19∗ .06 .55∗∗ .50∗∗

Note. obs. = observed.

∗
p < .05. ∗∗ p < .01.

and negative, over their students. Indeed, teachers seemed to ex- with students, students react against or compensate for this nega-
ert an invisible hand, as their provision of support and conflict was tivity by being more supportive of more of their peers.
associated with all of the three aspects of peer ecologies: richness Hierarchy in popularity was not associated with any of the
of interpersonal ties, status hierarchy, and social behaviors. teacher support or conflict measures in this study. The classroom
structure of popularity may be associated more with student and
group characteristics or other aspects of teaching than convey-
4.1. Richness of interpersonal ties
ing conflict and support. It could also be the case that by tak-
ing the classroom peer ecology as the unit of analysis, influen-
When a teacher showed support to more students, students re-
tial subgroups are overlooked. For instance, Moore, Shoulberg, and
ported more liking for each other, whereas when a teacher showed
Murray-Close (2012) found that teacher support and conflict af-
more conflict, students reported more disliking for each other.
fected popularity differently for boys and girls. For future research
These findings support the notion of student-specific teacher sup-
it seems necessary to delve deeper into the popularity issue, to
port and conflict as a social referent for peer relationships (Hughes
further unravel how the teacher might affect this type of status
et al., 2001; McAuliffe et al., 2009). However, contrary to our ex-
within the peer ecology.
pectations, student-specific teacher support was also positively re-
lated to peer disliking. This could be due to a teacher’s pet effect
4.3. Social behaviors
(Babad, 1995; 2009), when some students are seen by their peers
as the teacher’s favorite and as a result are disliked more. Impor-
Following our hypothesis, in classes with higher norms of
tantly, classes where the teacher has a pet who is not popular with
teacher support, students reported more prosocial behavior. This
the other students tend to have more negative classroom climate
may indicate that indeed a teacher’s stronger focus on positive
and lower student satisfaction and morale (Babad, 2009), which is
student characteristics models positive student interaction as well.
likely reflected by more peer disliking. Alternatively, this finding
Likewise, in classes with higher norms of teacher conflict, students
could be a consequence of students feeling secure in their rela-
reported more aggression, so teachers’ negative interactions may
tionship with the teacher and therefore feeling more freedom to
pave the way for negative student behavior. Of course, this asso-
indicate peers they dislike. Another possibility might be that in
ciation could also be due to the teacher having to intervene more
an overall positive classroom climate, a slightly negative feeling
often when students show aggressive behavior.
might result in a nomination for disliking sooner than in a class-
Unexpectedly, also differential teacher support was positively
room where negativity is to a larger extent the norm, due to a
associated with prosocial behavior. It could be the case that teach-
contrast effect. That is, the threshold (Terry, 2000) for indicating
ers deliberately differentiate in the amount of support in order to
dislike for anyone might be lower, which makes it difficult to ob-
accommodate students with specific needs. Teachers who do so
jectively compare the actual richness of negative ties among peer
may model for their students an inclination to help and cooper-
ecologies. For future research it would be worthwhile to addition-
ate with each other, that is, show prosocial behavior. Alternatively,
ally measure peer disliking and aggression from an outsider’s point
these results may indicate that in some situations peers compen-
of view.
sate for the negative consequences of a less equally supportive
teacher by making an effort to show more support to each other.
4.2. Status hierarchy For future research it might be interesting to investigate between-
class differences not only in the amount, but also in the nature of
As expected, in classrooms where the teacher more clearly con- teacher differential behavior.
centrated positive affect on specific students, more hierarchy ex-
isted in the peer ecology, but only for liking and not for popular- 4.4. Student-specific versus general teacher support and conflict
ity. The finding for hierarchy in liking status fits the notion of the
teacher as a social referent, that is, the teacher informed his class In line with the analyses by Den Brok et al. (2006) and Wubbels
on the differential value of the students. Surprisingly, the more et al. (2015) we conclude that the conceptualization of support and
teacher–student conflict was reported by students, the less hier- conflict the teacher expresses either towards specific students or
archical the distribution of likeability status was. This result may in more general terms deserves consideration. As expected, gen-
indicate that when the teacher has more conflicted relationships eral teacher support and the classroom norm of student-specific
38 M.M.H.G. Hendrickx et al. / Teaching and Teacher Education 53 (2016) 30–40

teacher support were positively associated, but correlations were, complex than they are depicted here. If the teacher and peers to-
at best, only moderate in strength (r range .19–.35). Therefore, gether form a system that evolves through time with elements mu-
these measures seem to be best interpreted as two distinct facets tually influencing each other (Sabol & Pianta, 2012), the peer ecol-
of teacher support. This was also reflected in our results for the ogy also affects teacher behavior. A classroom in which students
student-specific and general support measures. For future research all like each other and behave nicely probably makes it easier for
it is important to keep this distinction in mind and to conceptual- teachers to act in a supportive way, whereas in classrooms where
ize and operationalize classroom measures of teacher support and bullying and fighting are more common teachers may need to cor-
conflict according to the particular object of research. rect students more often. Although the plausibility of bidirectional
effects has been acknowledged by researchers (e.g., Bierman, 2011;
4.5. Observed and teacher-reported teacher support and conflict Gest & Rodkin, 2011), empirical research in this area is only in its
early stages. An important challenge for future research is to study
We did not find any associations between peer ecology and to what extent associations between teacher support and conflict
our observations or teacher self-reported support and conflict in on the one hand and peer ecology on the other are the result of
class. Possibly, the students’ perceptions of a supportive climate are an ongoing interaction between the two aspects of the classroom
more tightly connected to the peer ecology than the teacher’s or climate. Still, reciprocity in the associations between teacher prac-
an outsider’s perceptions. As described earlier, we expected that tices and peer ecology does not alter the possibility (or necessity)
the modeling and social referencing mechanisms imply that the for the teacher, as a responsible professional, to take action and
teacher affects the peer ecology through the students’ intake of exert her influence to guide the development of the peer ecology
what the teacher does. In the broader classroom climate literature in a desirable direction.
the importance of the students’ subjective perceptions and attribu- Secondly, although peer nominations have proven their value
tions has long been recognized (e.g., Fraser, 1989; Wubbels et al., extensively in prior studies (see Cillessen, 2009; Rubin et al., 2006),
2015). The general reasoning is that for individual student out- the peer ecology measures derived from them may provide a one-
comes, such as school engagement, general well-being, or external- sided image. For example, conclusions about social behaviors can
izing behavior, it might be more important whether a student feels only be derived from the students’ perceptions of each of their
supported by the teacher rather than exactly how a teacher treated peers, not about a more absolute level of prosocial behavior or ag-
the student. This is not to say that the teacher’s or an observer’s gression in the classroom climate in general. Directly asking par-
perception are irrelevant, but it may account for the finding that ticipants to rate the ecology as such could improve and enrich the
observations and the teacher’s perspective did not add to the ex- conceptualization of the peer ecologies (cf. Boor-Klip, Segers, Hen-
planation of the several peer ecology outcomes. drickx, & Cillessen, 2015).
For the observations in particular, the lack of significant results A final limitation is the sample size. Although most data en-
could be due to the fact that we were only able to record 2 h tries were based on information gathered from almost 1500 stu-
of video material per class. Therefore, it is possible that this ma- dents, data about our main focus of classroom peer ecologies and
terial was not representative, for example because we observed a teacher support regarded the 58 higher level cases. This clearly
teacher on a particularly positive or moody day. Another possibil- limited statistical power and only relatively strong effects could
ity is that the 2 h of video did not include rare, but influential be detected. Still, we found that some of the teacher support and
occurrences of teacher behavior. The students’ perceptions, how- conflict measures were significantly related to the peer ecology
ever, were based on a wealth of witnessed teacher–student inter- outcomes. Therefore, we encourage future research to investigate
actions, including those rare but potentially influential teacher in- these processes using larger samples of classrooms. Only then can
teractions. For instance, research on the negativity bias (e.g., Rozin the more subtle associations between the teacher and the class-
& Royzman, 2001; Vaish, Grossman, & Woodward, 2008) suggests room peer ecology be found and can the teacher’s touch be fully
that one negative interaction may have a far stronger impact than unraveled.
many positive ones. Thus, a single conflicted interaction between
the teacher and a student may have colored peer perceptions of 4.7. Implications for practice
this student’s level of teacher conflict, but was not recorded on
camera. Another possible explanation is that peer perceptions are The study findings indicate that for teachers, showing support
not only informed by observed teacher behavior, but are likely and avoiding conflict is important for the peer ecology within
to be biased by students’ prior judgments of their peers (Mikami which students interact with each other. Education and interven-
et al., 2012), and therefore may be more closely related to the peer tion programs focusing on classroom social dynamics might benefit
ecology than observed behavior. The positive correlations between from addressing how elevating the classroom level of teacher sup-
student-specific support measures and both observed interactions port may function as an effective intervention for altering the way
and student-perceived collective support, however, do suggest that students approach and value each other. In this sense, every-day
these nominations are at least partly associated with the teacher’s teacher practices could become network-related teaching strate-
actual interaction with students. Thus common-method variance gies (see Gest & Rodkin, 2011), when teachers deliberately apply
does at least not seem to be the only source for co-variation. them to affect the peer ecology. As an example, Reinke, Lewis-
Palmer, and Merrell (2008) studied an intervention called Class-
4.6. Limitations and future directions room Check-Up (CCU), in which they aimed to increase teach-
ers’ supportive comments and decrease their more conflicted, rep-
In sum, our study shows that in particular the level and distri- rimanding remarks. CCU did alter teacher behavior, and conse-
bution of student-specific teacher support and conflict is related to quently also students’ aggressive behavior declined.
several aspects of the peer ecology. These results need to be in- Building and maintaining supportive relationships with all stu-
terpreted in the light of some limitations. First, in this study we dents may not always be easy for teachers (Newberry, 2010). Emo-
have mainly explained the associations in terms of the teacher tionally or behaviorally demanding students place more relational
functioning as a model or social referent and as such influenc- press on teachers than those who are easily managed (Newberry &
ing the peer ecology. However, given the correlational design that Davis, 2008). Not reprimanding these students, but instead show-
was used, statements regarding causality cannot be made. Interre- ing support, might moreover interfere with classroom manage-
lations between the teacher and peer variables are probably more ment goals, such as preventing or decreasing disruptive behavior.
 M.M.H.G. Hendrickx et al. / Teaching and Teacher Education 53 (2016) 30–40 39

Following from our study, a solution might be not to reprimand Cillessen, A. H. N. (2009). Sociometric methods. In K. H. Rubin, W. M. Bukowski, &
students publicly for their behavior, but more privately. Because B. Laursen (Eds.), Handbook of peer interactions, relationships, and groups (pp. 82–
99). New York and London: The Guildford press.
in our study in particular the peer-perceived teacher support and Cillessen, A. H. N. (2011). Toward a theory of popularity. In A. H. N. Cillessen,
conflict measures were related to the peer ecology, teachers need D. Schwartz, & L. Mayeux (Eds.), Popularity in the peer system (pp. 273–299).
to be aware of supportive or conflictive interactions that are visi- New York, NY: Guilford.
Coie, J. D., Dodge, K. A., & Coppotelli, H. (1982). Dimensions and types of social
ble and/or audible to the other students in the classroom. So, the status: a cross-age perspective. Developmental Psychology, 18, 557–570.
combination of a private reprimand aimed at decreasing disruptive Cornelius-White, J. (2007). Learner-centered teacher-student relationships are effec-
behavior and a public expression of support another time might tive: a meta-analysis. Review of Educational Research, 77, 113–143. http://dx.doi.
org/10.3102/003465430298563.
serve both classroom management and peer ecology, and thus in-
Crick, N. R., & Grotpeter, J. K. (1995). Relational aggression, gender, and social-
dividual students’ development, best. psychological adjustment. Child Development, 66, 710–722.
Den Brok, P., Brekelmans, M., & Wubbels, T. (2004). Interpersonal teacher behaviour
and student outcomes. School Effectiveness and School Improvement, 15, 407–442.
5. Conclusion http://dx.doi.org/10.1080/09243450512331383262.
Den Brok, P., Brekelmans, M., & Wubbels, T. (2006). Multilevel issues in research
This study illustrates that in the classroom, teacher support using students’ perceptions of learning environments: the case of the ques-
tionnaire on teacher interaction. Learning Environment Research, 9, 199–213.
and conflict and peer ecologies are not isolated constructs but are http://dx.doi.org/10.1007/s10984-006-9013-9.
related to each other. Research aimed at investigating classroom Dishion, T. J., & Piehler, T. F. (2009). Deviant by design: peer contagion in develop-
social dynamics and student development, but also teachers and ment, interventions, and schools. In K. H. Rubin, W. M. Bukowski, & B. Laursen
(Eds.), Handbook of peer interactions, relationships, and groups (pp. 589–602).
teacher educators, thus should be cognizant of both constructs. Our New York and London: The Guildford press.
findings support the notion that the teacher functions as a model Doumen, S., Verschueren, K., Buyse, E., Germeijs, V., Luycks, K., & Soenens, B. (2008).
or social referent for students regarding how to interact and form Reciprocal relations between teacher-child conflict and aggressive behavior in
kindergarten: a three-wave longitudinal study. Journal of Clinical Child and Ado-
relationships with others. When teachers are aware of this, they
lescent Psychology, 37, 588–599. http://dx.doi.org/10.1080/15374410802148079.
can deliberately use their everyday interactions with students as Espelage, D. L., Holt, M. K., & Henkel, R. R. (2003). Examination of peer-group con-
network-related teaching strategies. Future research in this field textual effects on aggression during early adolescence. Child Development, 74,
205–220.
may build on these insights in further revealing the invisible hand
Farmer, T. W., McAuliffe Lines, M., & Hamm, J. V. (2011). Revealing the invisible
of the teacher. hand: the role of teachers in children’s peer experiences. Journal of Applied De-
velopmental Psychology, 32, 247–256. http://dx.doi.org/10.1016/j.appdev.2011.04.
006.
Acknowledgment Farmer, T. W., & Xie, H. (2007). Aggression and school social dynamics: the good,
the bad, and the ordinary. Journal of School Psychology, 45, 461–478. http://dx.
This research was supported by grant 411-10-915 from The doi.org/10.1016/j.jsp.2007.06.008.
Fraser, B. J. (1989). Twenty years of classroom climate work: progress and
Netherlands Organization for Scientific Research (NWO). prospect. Journal of Curriculum Studies, 21, 307–327. http://dx.doi.org/10.1080/
0022027890210402.
References Gest, S. D., & Rodkin, P. C. (2011). Teaching practices and elementary classroom peer
ecologies. Journal of Applied Developmental Psychology, 32, 288–296. http://dx.
Ahn, H. J., Rodkin, P. C., & Garandeau, C. F. (2010). Effects of classroom embedded- doi.org/10.1016/j.appdev.2011.02.004.
ness and density on the social status of aggressive and victimized children. Jour- Hodges, E. V. E., Boivin, M., Vitaro, F., & Bukowski, W. M. (1999). The power of
nal of Early Adolescence, 30(1), 76–101. friendship: Protection against an escalating cycle of peer victimization. Devel-
Babad, E. (1995). The “teacher’s pet” phenomenon, perceptions of teachers’ differ- opmental Psychology, 35, 94–101.
ential behavior, and students’ morale. Journal of Educational Psychology, 87, 361– Hughes, J. N., Cavell, T. A., & Willson, V. (2001). Further support for the develop-
374. mental significance of the quality of the teacher-student relationship. Journal of
Babad, E. (2009). The social psychology of the classroom. New York: Routledge. School Psychology, 39, 289–301.
Bierman, K. L. (2011). The promise and potential of studying the “invisible hand” of Hughes, J. N., & Chen, Q. (2011). Reciprocal effects of student-teacher and student-
teacher influence on peer relations and social outcomes: a commentary. Jour- peer relatedness: effects on academic self efficacy. Journal of Applied Develop-
nal of Applied Developmental Psychology, 32, 297–303. http://dx.doi.org/10.1016/j. mental Psychology, 32, 278–287. http://dx.doi.org/10.1016/j.appdev.2010.03.005.
appdev.2011.04.004. Hughes, J. N., Im, M. H., & Wehrly, S. E. (2014). Effect of peer nominations of
Boor-Klip, H. J., Segers, E., Hendrickx, M. M. H. G., & Cillessen, A. H. N. (2015). De- teacher–student support at individual and classroom levels on social and aca-
velopment and psychometric properties of the classroom peer context question- demic outcomes. Journal of School Psychology, 52, 309–322. http://dx.doi.org/10.
naire. Social Development Advance online publication. http://dx.doi.org/10.1111/ 1016/j.jsp.2013.12.004.
sode.12137. Hughes, J. N., Zhang, D., & Hill, C. R. (2006). Peer assessments of normative and
Borgatti, S. P., Everett, M. G., & Johnson, J. C. (2013). Analyzing social networks. Lon- individual teacher-student support predict social acceptance and engagement
don: SAGE Publications. among low-achieving children. Journal of School Psychology, 43, 447–463. http:
Bronfenbrenner, U. (1979). The ecology of human development: Experiments by nature //dx.doi.org/10.1016/j.jsp.2005.10.002.
and design. Cambridge, MA: Harvard University Press. Kindermann, T. A. (2007). Effects of naturally existing peer groups on changes in
Bronfenbrenner, U., & Morris, P. A. (2006). The bioecological model of human devel- academic engagement in a cohort of sixth graders. Child Development, 78(4),
opment. In R. M. Lerner, & W. Damon (Eds.), Theoretical models of human devel- 1186–1203.
opment: vol. 1. Handbook of child psychology (6th ed.) (pp. 793–828). Hoboken, Ladd, G. W., Birch, S. H., & Buhs, E. S. (1999). Children’s social and scholastic lives
NJ: John Wiley & Sons. in kindergarten: related spheres of influence? Child Development, 70(6), 1373–
Brown, B. B. (2011). Popularity in peer group perspective: the role of status in ado- 1400.
lescent peer systems. In A. H. N. Cillessen, D. Schwartz, & L. Mayeux (Eds.), Pop- Landis, J. R., & Koch, G. G. (1977). The measurement of observer agreement for cat-
ularity in the peer system (pp. 165–192). New York, NY: Guilford. egorical data. Biometrics, 33, 159–174. http://dx.doi.org/10.2307/2529310.
Buyse, E., Verschueren, K., Verachtert, P., & Van Damme, J. (2009). Predicting school Lapinski, M. K., & Rimal, R. N. (2005). An explication of social norms. Communication
adjustment in early elementary school: impact of teacher-child relationship Theory, 15, 127–147.
quality and relational classroom climate. The Elementary School Journal, 110, 119– Luckner, A. E., & Pianta, R. C. (2011). Teacher-student interactions in fifth grade
141. http://dx.doi.org/10.1086/605768. classrooms: relations with children’s peer behavior. Journal of Applied Develop-
Cappella, E., Kim, H. Y., Neal, J. W., & Jackson, D. R. (2013). Classroom peer re- mental Psychology, 32, 257–266. http://dx.doi.org/10.1016/j.appdev.2011.02.010.
lationships and behavioral engagement in elementary school: the role of so- Lüdtke, O., Robitzsch, A., Trautwein, U., & Kunter, M. (2009). Assessing the impact
cial network equity. American Journal of Community Psychology, 52, 367–379. of learning environments: how to use student ratings of classroom or school
http://dx.doi.org/10.1007/s10464-013-9603-5. characteristics in multilevel modeling. Contemporary Educational Psychology, 34,
Cappella, E., & Neal, J. (2012). A classmate at your side: teacher practices, peer vic- 120–131. http://dx.doi.org/10.1016/j.cedpsych.2008.12.001.
timization and network connections in urban schools. School Mental Health, 4, Mainhard, M. T., Brekelmans, M., & Wubbels, T. (2011). Coercive and supportive
81–94. http://dx.doi.org/10.1007/s12310-012-9072-2. teacher behaviour: within- and across-lesson associations with the classroom
Chang, L. (2003). Variable effects of children’s aggression, social withdrawal, and social climate. Learning and Instruction, 21, 345–354. http://dx.doi.org/10.1016/j.
prosocial leadership as functions of teacher beliefs and behaviors. Child Devel- learninstruc.2010.03.003.
opment, 74, 535–548. http://dx.doi.org/10.1111/1467-8624.7402014. Mantzicopoulos, P. (2005). Conflictual relationships between kindergarten children
Chang, L. (2004). The role of classroom norms in contextualizing the relations of and their teachers: associations with child and classroom context variables.
children’s social behaviors to peer acceptance. Developmental Psychology, 40, Journal of School Psychology, 43, 425–442. http://dx.doi.org/10.1016/j.jsp.2005.09.
691–702. http://dx.doi.org/10.1037/0012-1649.40.5.691. 004.
40 M.M.H.G. Hendrickx et al. / Teaching and Teacher Education 53 (2016) 30–40

McAuliffe, M. D., Hubbard, J. A., & Romano, L. J. (2009). The role of teacher cognition Runions, K. C. (2014). Reactive aggression and peer victimization from pre-
and behavior in children’s peer relations. Journal of Abnormal Child Psychology, kindergarten to first grade: accounting for hyperactivity and teacher-child con-
37, 665–677. http://dx.doi.org/10.1007/s10802-009-9305-5. flict. British Journal of Educational Psychology, 84, 537–555. http://dx.doi.org/10.
Mikami, A. Y., Griggs, M. S., Reuland, M. M., & Gregory, A. (2012). Teacher practices 1111/bjep.12037.
as predictors of children’s classroom social preference. Journal of School Psychol- Sabol, T. J., & Pianta, R. C. (2012). Recent trends in research on teacher-child re-
ogy, 50, 95–111. http://dx.doi.org/10.1016/j.jsp.2011.08.002. lationships. Attachment & Human Develoment, 14, 213–231. http://dx.doi.org/10.
Mikami, A. Y., Lerner, M. D., & Lun, J. (2010). Social context influences on children’s 1080/14616734.2012.672262.
rejection by their peers. Child Development Perspectives, 4, 123–130. http://dx. Schäfer, M., Kron, S., Brodbeck, F. C., Wolke, D., & Schulz, H. (2005). Bullying roles
doi.org/10.1111/j.1750-8606.2010.00130.x. in changing contexts: the stability of victim and bully roles from primary to
Moore, C. C., Shoulberg, E. K., & Murray-Close, D. (2012). The protective role of secondary school. International Journal of Behavioral Development, 29, 323–335.
teacher preference for at-risk children’s social status. Aggressive Behavior, 38, Statistics Netherlands (2012a). Aandeel allochtonen per gemeente [Percentages immi-
481–493. http://dx.doi.org/10.1002/ab.21446. grants by municipality] Retrieved from http://www.cbs.nl/nl-NL/menu/themas/
Muthén, B., & Muthén, L. (2014). Mplus (version 7.2) [computer software]. dossiers/allochtonen/cijfers/extra/aandeel-allochtonen.htm .
Neal, J. W., & Neal, Z. P. (2013). Nested or networked? Future directions for ecolog- Statistics Netherlands (2012b). Jaarrapport integratie 2012 [Annual report on integra-
ical systems theory. Social Development, 22, 722–737. http://dx.doi.org/10.1111/ tion 2012]. The Hague/Heerlen: Centraal Bureau voor de Statistiek.
sode.12018. Terry, R. (2000). Recent advances in measurement theory and the use of sociometric
Newberry, M. (2010). Identified phases in the building and maintaining of positive techniques. In A. H. N. Cillessen, & W. M. Bukowski (Eds.), Recent advances in the
teacher-student relationships. Teaching and Teacher Education, 26, 1695–1703. measurement of acceptance and rejection in the peer system. San Francisco, CA:
http://dx.doi.org/10.1016/j.tate.2010.06.022. Jossey–Bass.
Newberry, M., & Davis, H. A. (2008). The role of elementary teachers’ conceptions Thomas, D. E., & Bierman, K. L. the Conduct Problems Prevention Research Group.
of closeness to students on their differential behaviour in the classroom. Teach- (2006). The impact of classroom aggression on the development of aggressive
ing and Teacher Education, 24, 1965–1985. http://dx.doi.org/10.1016/j.tate.2008. behavior problems in children. Development and Psychopathology, 18, 471–487.
02.015. http://dx.doi.org/10.10170/S954579406060251.
Poulin, F., & Dishion, T. J. (2008). Methodological issues in the use of peer sociomet- Vaish, A., Grossman, T., & Woodward, A. (2008). Not all emotions are created equal:
ric nominations with middle child youth. Sociometric Assessment, 17, 908–922. the negativity bias in social-emotional development. Psychological Bulletin, 134,
http://dx.doi.org/10.1111/j.1467-9507.2008.00473.x. 383–403. http://dx.doi.org/10.1037/0033-2909.134.3.383.
Reinke, W. M., Lewis-Palmer, T., & Merrell, K. (2008). The classroom check-up: a Verschueren, K., Doumen, S., & Buyse, E. (2012). Relationships with mother, teacher,
classwide teacher consultation model for increasing praise and decreasing dis- and peers: unique and joint effects on young children’s self-concept. Attach-
ruptive behavior. School Psychology Review, 37, 315–332. ment & Human Development, 14, 233–248. http://dx.doi.org/10.1080/14616734.
Rodkin, P. C., & Hodges, E. V. E. (2003). Bullies and victims in the peer ecology: four 2012.672263.
questions for psychologists and school professionals. School Psychology Review, Wasserman, S., & Faust, K. (1994). Social network analysis: Methods and applications.
32, 384–400. Cambridge; New York: Cambridge University Press.
Roland, E., & Galloway, D. (2002). Classroom influences on bullying. Educational Re- Wentzel, K. R., & Caldwell, K. (1997). Friendships, peer acceptance, and group mem-
search, 44, 299–312. http://dx.doi.org/10.1080/0013188022000031597. bership: relations to academic achievement in middle school. Child Development,
Rosenthal, R., & Jacobson, L. (1968). Pygmalion in the classroom. New York: Holt, 68, 1198–1209.
Rinehart and Winston. Wubbels, T., Brekelmans, M., Den Brok, P., & van Tartwijk, J. (2006). An interpersonal
Roseth, C. J., Johnson, D. W., & Johnson, R. T. (2008). Promoting early adolescents’ perspective on classroom management in secondary classrooms in the Nether-
achievement and peer relationships: the effects of cooperative, competitive, and lands. In C. Evertson, & C. Weinstein (Eds.), Handbook of classroom management:
individualistic goal structures. Psychological Bulletin, 134, 223–246. http://dx.doi. Research, practice and contemporary issues (pp. 1161–1191). New York: Lawrence
org/10.1037/0033-2909.134.2.223. Erlbaum Associates.
Rozin, P., & Royzman, E. B. (2001). Negativity bias, negativity dominance, and con- Wubbels, T., Brekelmans, M., Den Brok, P., Wijsman, L., Mainhard, T., & Van
tagion. Personality and Social Psychology Review, 5, 296–320. http://dx.doi.org/10. Tartwijk, J. (2015). Teacher-student relationships and classroom management.
1207/S15327957PSPR0504_2. In E. T. Emmer, & E. J. Sabornie (Eds.), Handbook of classroom management (2nd
Rubin, K. H., Bukowski, W., & Parker, J. (2006). Peer interactions, relationships, and ed.) (pp. 363–386). New York and London: Routledge.
groups. In N. Eisenberg, W. Damon, & R. M. Lerner (Eds.), Social, emotional, and
personality development: vol. 3. Handbook of child psychology (6th ed.) (pp. 571–
645). Hoboken, NJ: John Wiley & Sons.