Cross-cultural variations in naïve psychology: a longitudinal comparison of preschool children in the United Kingdom and Singapore
This thesis presents a three-phase longitudinal study of naïve psychology and pretend play behaviour development between preschool children in the United Kingdom (UK) and Singapore. Research conducted in the Western contexts has shown that children develop an understanding of pretence and desires at 18 months of age (e.g. Nielsen & Dissanayake, 2004; Repacholi & Gopnik, 1997), before level-1 visual perspective-taking at 2½ years of age (e.g. Flavell, Everett, Croft, & Flavell, 1981) and followed by level-2 visual perspective-taking, appearance-reality distinction and false-belief understanding at 4 years of age (e.g. Flavell et al., 1981; Flavell, Flavell, & Green, 1983; Wellman, Cross, & Watson, 2001). A major issue that has dominated the field for many years concerns whether naïve psychology follows a universal developmental pattern. The majority of the studies to date have tended to rely heavily on false-belief understanding as an index of children’s understanding of mental representation. Some cross-cultural results have shown that the onset of false-belief understanding coincides with Western norms (e.g. Callaghan et al., 2005) whereas several non-Western studies have demonstrated a time lag in development across cultural groups (e.g. Vinden, 1999). To date no longitudinal study comparing the development of other naïve psychology concepts from 2 to 4 years of age between diverse cultures has been published. The present study aims to address the gap in the literature by tracking longitudinally and comparing the developmental patterns of children’s understanding of a range of naïve psychology concepts in the UK and Singapore at 2½, 3 and 3½ years of age (phases I, II and III respectively). Singapore with its mixed blend of Eastern and Western values represents a unique case for comparative study. This study employed a repeated-measures design, incorporating a large battery of established tasks that tapped children’s understanding of pretence, desires, visual perceptions and beliefs. In addition, a semi-structured observational approach was employed to study children’s naturally occurring pretend play behaviour. A total of 87 children were recruited in the UK (M = 28.60 months, SD = 1.90) and Singapore (M = 29.89, SD = 2.76) in the first phase of study. Of the initial sample, 36 children (M = 42.75, SD = 1.84) in the UK cohort and 38 children (M = 43.68, SD = 2.79) in the Singapore cohort participated in all three phases of the study. This thesis has five research questions. The first question relates to the extent to which acquisition of naïve psychology concepts differ between the two cultures at 2½ years of age. The baseline results reported in Chapter 5 indicate that 2½-year-old children in both cohorts acquired a rudimentary understanding of some aspects of pretence, discrepant desires, action prediction, emotion prediction and level-1 visual perspective-taking. The results showed no gross cross-cultural differences. However, subtle cross-cultural differences in children’s understanding of discrepant desires and action prediction were found. The second question addresses longitudinal cross-cultural differences in naïve psychology development between 2½, 3 and 3½ years of age. The results presented in Chapter 6 reveal cultural similarities in children’s performance on several pretence understanding, the level-2 visual perspective-taking, the appearance-reality distinction and the false-belief explanation tasks. Nonetheless, cultural differences were observed in some aspects of naïve psychology. The UK cohort performed significantly better than the Singapore cohort in the unexpected transfer false-belief prediction task at 3½ years of age, after verbal mental age (VMA) and gender were treated as covariates. Additionally, the UK cohort achieved significantly higher total mean for the level-1 visual perspective-taking task across the three phases and the mental representation in pretence task across phases II and III. In contrast, the Singapore cohort scored significantly higher in total mean for the discrepant desires task across the three phases. The third question considers longitudinal differences in children’s understanding of knowledge-ignorance and beliefs from 3 to 3½ years of age. The analysis in Chapter 7 indicates that the Singapore cohort performed significantly more poorly than the UK cohort in understanding knowledge-ignorance attribution (for the false-belief prediction and falsebelief explanation tasks) and true-belief ascription (for the false-belief explanation task) across phases II and III, after VMA and gender were considered as covariates. Comparison of children’s false-belief prediction and justification scores revealed that the cross-cultural difference in false-belief prediction related to an explicit ability to predict false-belief without concurrent ability to justify a naïve character’s behaviour based on false-beliefs. Twenty-four (66.7%) and 11 (28.9%) children in the UK and Singapore cohorts respectively were able to make correct false-belief prediction at 3½ years of age. Among these children, only six and five children from the UK and Singapore cohorts respectively provided correct justifications on the basis of false-beliefs. These findings therefore indicated cultural similarities in that the same number of children in both cohorts was able to predict and justify other’s behaviour in terms of false-beliefs. The fourth question explores the degree to which presence of sibling(s), birth order, language (VMA) and bilingualism contribute to individual differences in naïve psychology development. The results in Chapter 8 show no evidence that presence of sibling(s) and birth order facilitated understanding of action prediction, discrepant desires, level-1 visual perspective-taking, mental representation in pretence and false-belief prediction in either cohort. With respect to the role of language in children’s naïve psychology development, there were concurrent (within phase) associations between VMA and false-belief prediction at 3½ years of age and longitudinal associations between VMA at 2½ years of age and falsebelief prediction at 3½ years of age for both cohorts. These findings suggest that language ability contributes to individual differences in false-belief understanding. It is worth highlighting that not all aspects of naïve psychology and VMA were related. The fifth and final question focuses on longitudinal cross-cultural similarities and differences in pretend play behaviour and examines the links between pretend play behaviour and naïve psychology development. The observational data in Chapter 9 reveal that the Singaporean children spent significantly more time engaged in non-pretend play and non-social pretend play at 2½ years of age whereas the UK children spent significantly more time engaged in social pretend play. This finding contrasted with the marked cultural differences in naïve psychology development found at 3 and 3½ years of age. It is important to note that the UK and Singaporean children showed similar developmental sequences from non-pretend to non-social pretend and finally to social pretend play behaviour and from simple to complex forms of social pretend play behaviour. With respect to other pretend play behaviour, the UK children spent significantly more time engaged in positive complementary bids, negative conflict, other forms of pretence, metacommunication and in the pretend theme of outings, holiday and weather across all phases than the Singaporean children. The associations between some early pretend play behaviour and later acquisition of some naïve psychology concepts for both cultures provide partial support for the proposition that pretend play behaviour is an early marker of understanding mental representation. The reciprocal relationships between some pretend play behaviour and some naïve psychology concepts for the Singapore children alone provide partial support for the premise that pretend play behaviour and naïve psychology are closely related and intertwined. Taken together, the findings presented in this thesis extend our understanding of the gradual development of various naïve psychology concepts and pretend play behaviour between a Western and a hybrid culture. There were, however, substantial cross-cultural differences in the onset of some aspects of naïve psychology and pretend play behaviour. The roles of language, siblings and social pretend play behaviour in children’s naïve psychology development cannot be fully understood without considering culture as a frame of reference. The results of this study have a number of important implications for policy and practice including how pretend play should form an integral part of early childhood curriculum. Recommendations for further research are discussed.