THE INFLUENCE OF MATERNAL SENSITIVITY, SHYNESS AND LANGUAGE ABILITY ON PRESCHOOLERS’ PEER LIKABILITY CHEUNG HOI SHAN (B.Soc.Sci. (Hons), NUS) A THESIS SUBMITTED FOR THE DEGREE OF DOCTOR OF PHILOSOPHY DEPARTMENT OF PSYCHOLOGY NATIONAL UNIVERSITY OF SINGAPORE 2013 DECLARATION I hereby declare that this thesis is my original work and it has been written by me in its entirety. I have duly acknowledged all the sources of information which have been used in the thesis. This thesis has also not been submitted for any degree in any university previously. _________________________ Cheung Hoi Shan [13 February 2014] i Acknowledgements I am deeply grateful to the following persons and organisations: My supervisor A/P John Elliott, whose wisdom, warmth and care had kept me going in the past few years. Thank you for being such a great mentor! My thesis committee members, A/P Sim Tick Ngee, A/P Susan Rickard Liow and Dr Tan Seok Hui, for providing invaluable advice and guidance. Special thanks also go to Dr Tan, for generously sharing her lab space with me, and for inviting me to her lab meetings. Dr Iliana Magiati, for her valuable suggestions on how I could improve my thesis. A/P Melvin Yap, for his insightful comments on my study, and A/P Mike Cheung for his statistical training and advice. Professor Zeynep Biringen, for her patient guidance in the use of the EA Scales, and Ms Veronica Rivera for her help with the EA parallel coding. Staff of the Singapore Children’s Society, Ms Sue Cheng, Ms Chan Qingrong and Ms Rosie Lim, for help with the MBQS sort and for taking part in the EA coding at the early stage of this study. The Lee Foundation, the Singapore Children’s Society, the Tan Chin Tuan Foundation, the Faculty of Arts and Social Sciences and the Department of Psychology, for providing the research funds which made this study possible in the first place. My fieldwork team mates, Cheng Yi Hoong, Chu Yiu Tak, Low Pei Yee and Ruth Leong, for their help in data collection. Thanks for coming on board and sharing your insights with me. Ms Loo Bee Bee, Ms Loh Poh Yee, Ms Harlizah, Ms Norlela, Ms Janice Ow, Ms Irene Ho, Mr Soloman Seow and Ms Irene Nai, for the administrative and technical assistance. The Principals, teachers, parents and children who took part in my study. Thank you very much for supporting local research effort in this field! My family, for their love and unwavering support. ii Table of Contents Acknowledgements ................................................................................................. ii Summary ............................................................................................................... vii List of Tables ......................................................................................................... ix List of Figures ......................................................................................................... x List of Illustrations ................................................................................................. xi CHAPTER 1 INTRODUCTION .................................................................................................. 1 1.1 Overview ................................................................................................... 1 1.2 Theoretical Framework – Attachment Theory .......................................... 3 1.2.1 Ethological Research and Bowlby’s Contributions ........................ 3 1.2.2 The Sensitivity Hypothesis.............................................................. 5 1.2.3 Emotional Availability .................................................................... 9 1.2.4 Maternal Sensitivity and Children’s Social Development ............ 13 1.2.5 Attachment and Sensitivity – Cross-cultural Considerations ........ 17 1.3 Child Variables – Shyness and Language Skills ..................................... 22 1.3.1 Temperament in Child Development ............................................ 22 1.3.2 The Conditional Model and Differential Susceptibility Hypothesis ................................................................................................................ 25 1.3.3 Shyness and Children’s Social Development: The Role of Culture and Gender ............................................................................................. 27 1.3.4 Language Ability and Children’s Social Development ................. 32 1.4 Associations Among the Predictors ......................................................... 34 1.4.1 Mediating Role of Language Ability ............................................ 36 1.5 Children’s Likability by Peers ................................................................. 38 iii 1.5.1 Conceptualisations and Measures of Children’s Likability Within the Peer Group ........................................................................................ 39 1.5.2 Sociometric Nominations and Peer ratings – A Comparison ........ 42 1.5.3 Teacher’s Ratings on Peer Likability ............................................ 43 1.5.4 Ethical Issues and Participation Rates in Sociometric Measures .. 45 1.6 Contributions of the Present Study .......................................................... 47 1.7 Hypotheses............................................................................................... 48 CHAPTER 2 METHODOLOGY ............................................................................................... 51 2.1 Participants .............................................................................................. 51 2.2 Measures .................................................................................................. 52 2.2.1 Maternal sensitivity ....................................................................... 52 2.2.2 Child Shyness ................................................................................ 56 2.2.3 Children’s Language Ability ......................................................... 57 2.2.4 Peer Likability ............................................................................... 61 2.2.5 Non-verbal Ability ........................................................................ 62 2.3 Procedure ................................................................................................. 63 2.4 Planned Statistical Analyses .................................................................... 66 2.5 Pilot Study ............................................................................................... 67 2.5.1 Participants .................................................................................... 69 2.5.2 Measures ........................................................................................ 69 2.5.3 Procedure ....................................................................................... 70 2.5.4 Results and Discussion of the Pilot Study ..................................... 72 iv CHAPTER 3 RESULTS ............................................................................................................. 75 3.1 Preliminary Analyses ............................................................................... 75 3.1.1 Screening for Missing Data, Outliers, and Multicollinearity ........ 75 3.1.2 Evaluation of Observation Sessions .............................................. 77 3.1.3 Internal Consistency ...................................................................... 77 3.1.4 Inter-observer Reliability .............................................................. 79 3.1.5 Cross-sex and Same-sex Peer Ratings .......................................... 81 3.1.6 The Influence of Participation Rate on Peer Ratings .................... 82 3.1.7 Screening for Covariates ............................................................... 84 3.2 Main Analyses ......................................................................................... 85 3.2.1 Bivariate Correlations ................................................................... 85 3.2.2 Hypotheses 1 and 2 - Predictors of Peer Likability ....................... 87 3.2.3 Hypothesis 3 – Gender as a Moderator ......................................... 92 3.2.4 Hypothesis 4 – Test of the Differential Susceptibility Hypothesis 93 3.2.5 Hypothesis 5 – Vocabulary Skills and Pragmatic Language Difficulty as Mediators ........................................................................... 94 3.3 Miscellaneous Analysis ........................................................................... 98 CHAPTER 4 DISCUSSION ..................................................................................................... 101 4.1 Summary of Findings ............................................................................ 101 4.2 Peer Likability – Gender and Cross-informant Differences .................. 102 4.3 Predictors of Peer Likability .................................................................. 103 4.4 Differential Susceptibility and the Mediating Effect of Language Ability……………………………………………………………………...105 v 4.5 Culture-specificity of the Values of Child Shyness and Maternal Sensitivity .................................................................................................... 107 4.6 Strengths and Limitations of the Study ................................................. 116 4.7 Practical Implications of the Research Findings ................................... 119 4.8 Directions for Future Research .............................................................. 120 4.9 Conclusion ............................................................................................. 123 References ........................................................................................................... 124 Appendix A Descriptions of the Five Dimensions of the Emotional Availability (EA) Scales ............................................................................................................................. 152 Appendix B Questionnaires Administered Participants’ Background Data .................................................................... 155 Mother’s Evaluation of the Play Session ..................................................... 157 Children’s Behaviour Questionnaire ........................................................... 158 Language Background Questionnaire ......................................................... 160 Relationship between a caregiver and child ................................................ 162 Filmer’s Evaluation of the Play Session ...................................................... 167 Appendix C Stimuli Used in Free Play Session and Peer Rating Procedure C1: Toys used in the 30-minute free play session ....................................... 168 C2: Pictures for demonstrating the peer rating procedure ........................... 169 C3: Pictures for children’s familiarisation with the peer rating procedure . 170 C4: Faces used for the 3-point peer likability rating scale .......................... 171 vi Summary Background and Aims Research has shown that having good relationships with peers is advantageous to children’s well-being, making it relevant to examine the factors which may predict children’s likability in the peer group. This study looked at how maternal sensitivity, child shyness, children’s vocabulary and pragmatic language skills may predict likability by peers in preschool. Past studies have shown how these predictors were individually related to children’s quality of peer relationships. This dissertation extends existing research by examining in a single study the relative contributions made by these four variables in the prediction of children’s likability by peers. Of particular interest is how maternal sensitivity and child shyness may be linked to likability in the local context, with implications for the cross-cultural relevance of these concepts in Singapore. The differential susceptibility hypothesis, which posits that maternal sensitivity interacts with child shyness in influencing children’s likability by peers, was also tested; along with language skills as mediators in the associations between sensitivity and shyness on the one hand, and likability by peers on the other. Method The sample included 164 mother-child dyads belonging to the four major ethnic groups in Singapore (Chinese, Malay, Indian and Others). Children (72 boys, 92 girls) were between ages 52 and 79 months (M = 67.07, SD = 6.67). Mother-child interactions in a free play session were coded using the Emotional Availability (EA) Scales (fourth edition) as a measure of maternal sensitivity. Mothers also completed measures on child shyness (Children’s Behaviour Questionnaire) and pragmatic language difficulty (Children’s Communication Checklist-2). Receptive and expressive vocabularies were tested using the Expressive One-word Vocabulary Test-4 and the locally-developed Bilingual vii Language Assessment Battery respectively. In the third quarter of the school year, the author visited the preschools to obtain likability ratings from classmates. Results and Discussion Gender differences were found in the association between the predictors and likability by peers. Teachers and children also seemed to judge children’s likability by peers on different bases. Shyness in boys was related to their being more well-liked by female peers, implying that shyness has a positive connotation within the opposite sex peer group in this culture. Contrary to expectation, maternal sensitivity was related to boys’ being less well-liked by girls. Low pragmatic language difficulty was associated with girls’ likability by male peers, with it being the only significant predictor. Furthermore, better vocabulary skills predicted more favourable teachers’ ratings of peer likability for boys; while low shyness and pragmatic language difficulty predicted the same for girls. These findings vindicated the use of a cross-informant methodology, in order to provide a more complete picture of children’s relative standing within the peer group. The differential susceptibility hypothesis was, however, not supported. Vocabulary skills and pragmatic language difficulty mediated the associations between maternal sensitivity, child shyness and peer likability as rated by teachers, but only for girls. This suggests the importance of language ability in accounting for variations in how well-liked children were by peers. The results and details of the method raised questions as to the importance of maternal sensitivity in Singapore families, and as to whether or not the EA Scales validly measured it, since the results are not entirely consistent with those from studies elsewhere. This dissertation, being the first local study conducted to look at how maternal sensitivity (as understood in the western context) was related to peer likability, has brought to light possible cultural variations in the understanding of the concept of sensitivity. Local research is very much needed to further understand how best sensitivity may be measured in Singapore. viii List of Tables Table 1: Bivariate correlations between sensitivity, shyness and peer acceptance ......................................................................................................... 72 Table 2: EA Scales scores provided by a native and non-native coder … ....... 81 Table 3: Bivariate correlations of the variables as a function of different participation rates ............................................................................................. 83 Table 4: Means and standard deviations, and bivariate correlations among the variables, differentiated by gender. .................................................................. 86 Table 5: Unique variances contributed by all the predictors towards boys’ likability by peers ............................................................................................. 89 Table 6: Unique variances contributed by all the predictors towards girls’ likability by peers ............................................................................................. 90 Table 7: Main and moderation effects for gender, maternal sensitivity and shyness .............................................................................................................. 93 Table 8: Vocabulary and pragmatic language difficulty as mediators in the relationship between maternal sensitivity and teachers’ ratings on peer likability ............................................................................................................ 97 Table 9: ICCs for inter-observer reliability of the other five EA dimensions .. 99 Table 10: Means and standard deviations, and bivariate correlations between the EA dimensions and the other variables .................................................... 100 Table 11: Sub-scales of the other five EA dimensions ................................... 154 ix