The Impact of the Classroom Peer Context - Publicatie-online (PDF)

T h e I m p a c t o f t h e C l a s s r o o m P e e r C o n t e x t o n C h i l d r e n ’s P e e r R e l a t i o n s h i p s H e n r i k e B o o r - K l i p T H E I M PA C T O F T H E C L A S S R O O M P E E R C O N T E X T O N C H I L D R E N ’ S P E E R R E L AT I O N S H I P S Henrike Boor-Klip 20 17 Uitnodiging Voor het bijwonen van de openbare verdediging van mijn proefschrift The impact of the classroom peer context on children’s peer relationships Op maandag 15 mei 2017 om 16.30u. In de aula van de Radboud Universiteit, Comeniuslaan 2, Nijmegen Receptie na afloop Henrike Boor-Klip [email protected] Paranimfen Marloes Hendrickx Roelinda Klip 14527-klip-cover.indd 1 27/03/2017 16:01   The Impact of the Classroom Peer Context on Children’s Peer Relationships Henrike Boor-Klip 14527-klip-layout.indd 1 27/03/2017 17:21  ISBN: 978-94-6299-567-3 Printing: Ridderprint BV - www.ridderprint.nl Cover and lay-out: Design Your Thesis - www.designyourthesis.com The research in this dissertation was funded by the Dutch Programme Council for Educational Research [NWO-PROO Grant No. 411-10-915]. © Henrike Boor-Klip, 2017 All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any means, electronical or mechanical, including photocopy, recording or any information storage or retrieval system, without the prior written permission of the copyright owner. 14527-klip-layout.indd 2 27/03/2017 17:21  The Impact of the Classroom Peer Context on Children’s Peer Relationships Proefschrift ter verkrijging van de graad van doctor aan de Radboud Universiteit Nijmegen op gezag van de rector magnificus prof. dr. J.H.J.M. van Krieken, volgens besluit van het college van decanen in het openbaar te verdedigen op maandag 15 mei 2017 om 16.30 uur precies door Hendrika Johanna Klip geboren op 21 augustus 1986 te Harderwijk 14527-klip-layout.indd 3 27/03/2017 17:21  Promotoren Prof. dr. A.H.N. Cillessen Prof. dr. P.C.J. Segers Prof. dr. J.M.G. Brekelmans (Universiteit Utrecht) Manuscriptcommissie Prof. dr. P.C. Meijer (voorzitter) Prof. dr. C.J. Rieffe (Universiteit Leiden) Dr. C.F. Garandeau (Universiteit Utrecht) 14527-klip-layout.indd 4 27/03/2017 17:21  CONTENTS Chapter 1 General Introduction 7 Chapter 2 Development and Psychometric Properties of the Classroom Peer Context Questionnaire 21 Chapter 3 Perceptions of Classroom Peer Context: Associations with Social Status, Academic Achievement, and Self-concept 45 Chapter 4 A Positive View of the Peer Context Moderates the Association between Low Status and Poor School Adjustment 63 Chapter 5 The Moderating Role of Classroom Descriptive Norms in the Association of Student Behavior with Social Preference and Popularity 89 Chapter 6 Guiding the Invisible Hand: Improving Classroom Peer Relationships through Tools for Teachers 115 Chapter 7 General Discussion 141 Appendices 155 References 163 Nederlandse Samenvatting 179 Publications 189 Curriculum Vitae 195 Dankwoord 199 14527-klip-layout.indd 5 27/03/2017 17:21  14527-klip-layout.indd 6 27/03/2017 17:21  CHAPTER 1 General Introduction 14527-klip-layout.indd 7 27/03/2017 17:21  14527-klip-layout.indd 8 27/03/2017 17:21  GENERAL INTRODUCTION | 9 1 Peer relationships are important in children’s development. This is demonstrated by decades of research showing that the nature and quality of peer relationships are associated with many social and academic outcomes (see, for an overview, Rubin, Bukowski, & Laursen, 2009). Peer relationships can be studied at different levels. Typically, three levels are distinguished: individual, dyad, and group (see, e.g., Rubin et al., 2009). At the individual level, the focus is on child behavior in interactions with peers as well as the position of a child among his or her peers. At the dyadic level, relationships among two children (e.g., friendships and antipathies) are studied. At the group level, patterns of interactions and relationships among three or more children (e.g., a classroom) are of interest. In this thesis, children’s peer relationships in school were examined at the first and third levels. At the individual level, I studied children’s social status (social preference and popularity) as well as individual social behaviors in interactions with peers. At the group level, I examined the classroom peer context or the overall nature of the peer group and peer interactions in the classroom as a whole. In previous peer relationships research, the individual level has received by far the most empirical attention. Contrary, the group level, and especially the classroom peer context, has received relatively little empirical attention, despite its theoretical importance (see, e.g., Bronfenbrenner, 1979; Gronlund, 1959). Therefore, the aim of this thesis was to increase our understanding of the role of the classroom peer context in children’s peer relationships by addressing three topics. First, I examined how the classroom peer context can be measured with a special focus on children’s own perceptions of this context. Second, I studied whether the classroom peer context moderates peer processes at the individual level. Third, I examined how teachers may improve peer relationships in the classroom as a whole. Children’s Social Status among Peers Social status is a central construct in peer relationships research, examining peer relationships at the individual level. Social status reflects children’s position among their peers and can be seen as an indicator of their social competence (Rose-Krasnor, 1997). From middle childhood on, two types of social status can be distinguished: social preference (or likeability) and popularity (Cillessen & Marks, 2011; Parkhurst & Hopmeyer, 1998; van den Berg, Burk, & Cillessen, 2015). Preference refers to the extent to which children are liked by their peers. Popularity refers to the visibility and power children have in their classroom. A large literature since the 1980s has described the causes of individual peer social status. Although researchers also have focused on the social-cognitive and 14527-klip-layout.indd 9 27/03/2017 17:21  10 | CHAPTER 1 emotional determinants of peer status, the main focus of this literature has been on social behaviors as the determinants of social status in the peer group (see, e.g., Coie, Dodge, & Kupersmidt, 1990; Rubin et al., 2009). Children’s social status largely depends on their behaviors in interactions with others (e.g., Cillessen & Mayeux, 2004; van den Berg et al., 2015). Children who show much prosocial behavior and little aggression or social withdrawal tend to be liked by their peers (Asher & McDonald, 2009; Lease, Musgrove, & Axelrod, 2002; Newcomb, Bukowski, & Pattee, 1993). Popular children also show frequent prosocial behavior and little social withdrawal (Asher & McDonald, 2009; LaFontana & Cillessen, 2002; Lease et al., 2002). However, contrary to well-liked children, popular children may also use aggression (Rose, Swenson, & Waller, 2004; Vaillancourt & Hymel, 2006). In addition to the causes and determinants of social status, researchers and practitioners alike also have been concerned about the consequences and outcomes associated with peer status. Peer rejection has been associated with a range of negative short-termandlong-termconsequencesincludinginternalizingandexternalizingproblems (van Lier & Koot, 2010), low self-concept (Spilt, van Lier, Leflot, Onghena, & Coplin, 2014), loneliness (Newman Kingery, Erdley, & Marshall, 2011), low academic achievement (e.g., Bellmore, 2011; Newman Kingery et al., 2011), and drop-out (Parker & Asher, 1987). Less is known about the consequences of popularity, yet being unpopular as well as being popular have been found to be related to lower academic achievement (e.g., Bellmore, 2011; Schwartz, Gorman, Nakamoto, & McKay, 2006; Troop-Gordon, Visconti, & Kuntz, 2011) and more health risk behaviors (e.g., Prinstein, Choukas-Bradley, Helms, Brechwald, & Rancourt, 2011). To summarize, children’s peer relationships at the individual level can be described as a two-step process going from behaviors to peer group social status and from peer group social status to clinical and adjustment outcomes. This two-step process describes how status is acquired over time and what consequences are associated with status over the course of a school year or even the longer term, once status is acquired. Of course, feedback loops between the elements of these steps are also possible (e.g., negative adjustment outcomes may further decrease a child’s peer acceptance). The Context in Peer Relationships Research When the causes and consequences of peer relations have been studied in the literature, it has often been assumed that they are the same across classrooms, schools, and other (broader) contexts. It is only in recent years that contextual effects (such as effects of 14527-klip-layout.indd 10 27/03/2017 17:21  GENERAL INTRODUCTION | 11 1 the group) on peer relationship processes have started to receive more attention both conceptually and empirically (e.g., Becker & Luthar, 2007; Mikami, Lerner, & Lun, 2010). This attention to context is inspired by Bronfenbrenner’s ecological systems theory (1979; see also Bronfenbrenner & Morris, 2006). In his theory, Bronfenbrenner poses that development takes place through processes that are interactions between an individual and the immediate environment. An example of such a process is the interaction between a child and her or his peers. However, Bronfenbrenner stressed that these processes are not universal, but depend, among others, on the immediate and remote context in which they occur. Therefore, it is also likely that the associations among behavior, social status, and social and academic outcomes (i.e., school adjustment) depend on the context in which they occur. One very important social context for peer relationships is the classroom. In fact, this is historically so obvious (see, e.g., Gronlund, 1959), that is it surprising that not more attention has been given in the literature to classroom context effects. The classroom is an important context for peer processes, because children spend a large part of their day in the classroom with a fixed group of peers (Ryan & Ladd, 2012). Each classroom has a unique peer context or peer ecology (Gest & Rodkin, 2011). This classroom peer context emerges from the interactions and relationships among the individual classmates (Rubin, Bukowski, & Parker, 2006). Although peer relationships have often been studied in classrooms, the effects of the classroom peer context have remained underexposed in the empirical literature as researchers typically have standardized data within classrooms (Cillessen & van den Berg, 2012; Rubin et al., 2006). However, this is changing. Several recent studies have considered the effects of the classroom peer context and have demonstrated its relevance for children’s peer relationships (e.g., Ahn, Garandeau, & Rodkin, 2010; Chang, 2004; Meisinger, Blake, Lease, Palardy, & Olejnik, 2007; Mikami et al., 2010; Serdiouk, Rodkin, Madill, Logis, & Gest, 2015). The current thesis makes a further and unique contribution to these studies by focusing on the overall peer relationships and interactions in a classroom at the end of primary school. The classroom peer context is a broad construct that includes, among others, the level of competitiveness or cooperativeness in the classroom, the extent to which children are friends, and the degree of unity or cohesion in the classroom (i.e., is the class a social group or are children pretty much on their own). Variation in these characteristics may have a large impact on how children function in school. I studied this impact by addressing the measurement of the classroom peer context, its effects on classroom peer relationships in primary school, and the ways in which teachers may change the classroom peer context. 14527-klip-layout.indd 11 27/03/2017 17:21  12 | CHAPTER 1 Topic 1: Measuring Classroom Peer Context In order to fully understand the impact of the classroom peer context on children’s peer relationships, it is important to first consider how it can be measured. Previous studies have used a variety of methods to assess classroom context. Most studies that examined peer relationships of older children and adolescents applied either sociometric methods (peer nominations or ratings) or social network analyses (Cillessen, 2009; Kindermann & Gest, 2009). For example, researchers interested in classroom aggression norms have asked children to nominate peers who are aggressive and computed a classroom norm from these nominations (e.g., Chang, 2004; Stormshak et al., 1999). Researchers interested in the social structure of the classroom asked children to report who hangs out with whom and used social network analyses to determine density and embeddedness of the network (e.g., Ahn et al., 2010). Less frequently used approaches to assess the classroom peer context are observations by researchers (Fabes, Martin, & Hanish, 2009) and teacher ratings (Gest, 2006). As examples, Frey, Higheagle Strong, and Onyewuenyi (2016) observed children’s behavior on the playground to examine classroom rates of aggression, while Gest (2006) used teacher ratings to examine the social structure of the classroom (i.e., friendships and social groups). Although each of these approaches offers a unique perspective on the classroom peer context and therefore has its value, it is not surprising that most peer relationship researchers rely on children as reporters of the classroom peer context. Children are likely to have the most complete picture of their classroom peer relationships. Observations by researchers usually happen only over a short period of time and teachers are also not always aware of all interactions among the children in their classroom. Despite the fact that sociometric and social network methods have both relied on children as informants to assess aspects of the classroom peer context (e.g., Ahn et al., 2010; Chang, 2004), a disadvantage of these methods in studying the classroom peer context is that researchers do not actually ask the children how they view their classroom. That is, the children are not aware of the exact question of the researchers nor answer it directly. For example, when peer nominations are used to examine the classroom level of aggression, researchers ask children to nominate peers who his, kick or push other children. From this question, children may derive that the researcher is interested in aggression in the classroom, but they do not know the exact question the researcher is interested in. Consequently, it could be that children experience their classroom as more (or less) aggressive than the researcher assumes based on the peer nominations. Because children may have the best insight in the classroom peer context, it is important to ask them for their own direct views of the classroom peer context. How 14527-klip-layout.indd 12 27/03/2017 17:21  GENERAL INTRODUCTION | 13 1 children themselves view the peer context is not captured with the previously described methods. Self-report measures do capture this perspective. Self-reports are often described as measures that have disadvantages (see Stone et al., 2000). Yet, in this case when the child’s own perspective is of interest, self-reported views are critical to know. This does not mean that self-reported measures should replace other measures of the classroom peer context. Instead, they should be used in addition to other measures of the classroom peer context. There are a few existing measures for examining the peer context in primary school, but they do not assess the peer context systematically or in-depth. Some measures only include a single scale (e.g., Brock, Nishida, Chiong, Grimm, & Rimm-Kaufman, 2008; Rowe, Kim, Baker, Kamphaus, & Horne, 2010). Others have multiple scales but focus only on negative aspects of the classroom (Fisher & Fraser, 1981) or only on interactions and not on relationships (e.g., Donkers & Vermulst, 2011). Therefore, these measures do not provide a comprehensive view of children’s perceptions. In order to get a more complete understanding of the classroom peer context from the children’s point of view, we developed a new measure that included multiple scales focusing on both positive and negative interactions and relationships in the peer context: the Classroom Peer Context Questionnaire (CPCQ). The development of this instrument and its psychometric properties are described in Study 1 (Chapter 2). Ideally, there is some level of agreement about the quality of the classroom peer context among children in the same classroom. This would confirm that there is actually a classroom peer context that can reliably be observed by the children (see Marsh et al., 2012) and that children’s perceptions are not all subjective or idiosyncratic. The more consensus there is among the children in a classroom, the more it can be trusted that these perspectives are a “real” description of what is actually going on in the classroom. Yet, it seems unlikely that all children perceive the classroom peer context in exactly the same way. Not all children have the same experiences with their classroom peers (Ladd & Troop-Gordon, 2003). Also, some children may just think more positively about the world around them than others (Baumeister, Campbell, Krueger, & Vohs, 2003). In Study 2 (Chapter 3), it is examined how individual differences between children were related to their perceptions of the classroom peer context. When it comes to the measurement of the classroom peer context, there is a variety of methods that can be considered. Some methods, including self-report measures, present an additional methodological choice in the operationalization of the peer context as both individual scores and group scores can be derived from them. In the case of individual scores, all children have their own score for the classroom peer context. In the case of group consensus scores, the same score is used for all children in 14527-klip-layout.indd 13 27/03/2017 17:21