Are popular children more likely than unpopular children to have developed a representational theory of mind at age 4? Sue Dockett and Sheila Degotardi Faculty of Education University of Western Sydney, Macarthur Paper presented at AARE, November, 1995. Abstract Recently, within the field of early childhood education, there has been an increasing emphasis on the role of social construction of knowledge and on the inter-relatedness of aspects of social and cognitive development. This paper will report on an exploratory study which investigates a proposed relationship between young children's popularity status among peers and a representational theory of mind. Based on a series of interviews, a group of four-year-old children were asked to rate their peers by posting photographs into boxes marked with a happy face, a sad face or a neutral face. These ratings were collated to generate an overall likability measure, which was then compared with a measure of performance on a series of false belief, appearance-reality and representational change tasks. Comparison of these measures indicated a significant correlation between the ranking of individual children as popular, or unpopular, by their peers and a measure of theory of mind. This connection is explained by reference to the view that social interactions among peers provide opportunities for understanding self and others and for explaining and predicting the actions of others based on mental states. Implications for early childhood education programs and possible avenues for future research are considered. Introduction During the last decade, considerable research attention has been paid to investigating young children's developing theories of mind. This research has focussed on describing the developments that occur within the early childhood years as well as identifying and investigating situations in which children demonstrate an understanding several aspects of this development, such as the appearance-reality distinction, false-belief and representational change. Children's theories of mind, in the current research context, are described as understandings children have of their own minds and others' minds and of the relation between the mind and the world. This understanding enables children to predict and explain actions by ascribing mental states, such as beliefs, desires and intentions, to themselves and to other people (Astington 1991, p. 158) Significant progress is made in understanding these areas between the ages of about four and six years. The basis for this progress is reported to be the ability to create and work with at least two conflicting mental representations in relation to the same object or person. This, in turn, is reliant on an understanding of the mind as an interpretive mechanism, rather than as a mechanism that faithfully reproduces the physical world (Perner, 1991). Recent research (Dockett 1994) investigated a proposed connection between children's involvement in complex sociodramatic play and their developing theories of mind. As a result of this investigation, it was noted that children who had engaged in a period of play intervention, where the nature and type of play was complex and elaborate, performed at a significantly improved level on standard theory of mind tasks when compared to a control group who had not engaged in the same intervention. One conclusion from this research was that the social situations in which children engage have an impact on the development that occurs. This research also raised the possibility that the children for whom the most dramatic increases in performance were observed were the children who were regarded as popular by their peers, and who were readily invited and accepted into a variety of social situations. In contrast, it was suggested that children who were unpopular with their peers may have been excluded from complex play situations and so were not engaged in the social situations which were identified as promoting, as well as reflecting, cognitive change. The aim of this research was to investigate the proposed connection between young children's popularity status and their performance on a series of theory of mind tasks. Such a project has been mentioned in international theory of mind discussions, (for example at the 1991 Society for Research in Child Development Conference), but has not been pursued in depth. While the aim of the research was not to try to establish a cause and effect relationship between these two areas, this may well be an area for future study. Recent research in the area of social development has highlighted the idea that children's friendships and peer acceptance have several implications for development in general, and social development in particular. Being accepted as a preschooler has been linked to the development of social competence (Kemple, 1991), and conversely, being rejected as a preschooler has been linked to a recursive cycle of rejection and isolation that persists into later years (Katz, 1988). In contrast to earlier research, Dunn (1993) has reported that young children can and do develop friendships which have a substantial impact on their interactions and their learning. Howes (1988) has reported that preschoolers are able to maintain friendships for an average of two years and Dunn's work with four- and five-year-olds has supported the description of such relationships as important in the lives of young children. Children have been reported to interact in different ways with friends and non-friends. For example, children spend more time with their friends (Hartup 1989) and they react differently in conflict situations by attempting conciliation rather than confrontation (Dunn, 1993). Dunn also reports that children may actually spend more time arguing with their friends, but that they resort to conflict less often in an attempt to resolve such arguments. When children interact with their friends they are more likely to engage in complex interactions, such as play, where some of the 'rules' for interaction and some of the themes or scripts to be enacted have already been established. In relation to theory of mind research, there is an increasing focus on the social context and the impact of this on cognitive development. One criticism of Piagetian theory has been the focus on the individual and the individual nature of cognitive change. Recently, there has been increased interest in the social context of cognitive development, with the ideas of Vygotsky being revisited and reconsidered. As a result, there is a growing body of research which indicates that the social context is instrumental in generating and promoting cognitive change (Rogoff, 1990). In this study, it is suggested that the social context of the peer group, or friendship group, plays an important role in cognitive development in at least two ways. Firstly, through the development of intersubjectivity, it is proposed that children establish a shared focus within their play and interactions, which in turn, provides the basis for communication about that focus, resulting in shared problem-solving and decision making that is then internalised by the individual participants. In essence, this view predicts that children will be able to achieve a great deal more in a group which has a shared focus than they otherwise would on their own. Secondly, the Piagetian view predicts that social interactions provide the impetus for cognitive development as children test out their understandings. When confronted with views, perspectives or understandings which differ from their own, children are described as seeking to resolve the cognitive conflict. Regardless of the proposed reasons for the importance of complex and sustained social interactions, children who are excluded regularly from such interactions would be expected to have fewer opportunities to demonstrate or to construct understanding of themselves and others. In summary, this research aims to investigate the hypothesis that popular children are more likely than unpopular children to have developed a representational theory of mind at age 4. This research is exploratory in that it is seeking to establish whether or not a connection between two areas exists. Methodology Sample The sample comprised one group of 24 children attending the same preschool in metropolitan south-western Sydney. All children had attended the same full-day preschool since the beginning of the year and participated in the data collection phase during terms 2 and 3, 1995. The average age of the thirteen girls and eleven boys was 54.75 months, with the range being 45 months to 64 months. Procedures The study was collected in three phases: ¥orientation visits ¥interviews to determine peer popularity ¥theory of mind interviews The first of these phases consisted of a series of orientation visits made to the centre by the research team. The aim of these visits was to establish a level of familiarity between the researchers and the children and to take the photographs of individual children which would be used in the second phase of the study. The second phase involved an interview between the research assistant and individual children. Using the procedure developed by Asher, Singleton, Tinsley and Hymel (1979) and modified during recent research (Denham, McKinley, Couchoud & Holt, 1990), children were shown photographs of their classmates (those taken during the orientation visits) and asked to name them and then to 'post' each photo into a box on which there were drawings of positive, neutral and sad faces. Before placing the photos into the boxes, children were asked to consider how much they liked the child in the photo. They were asked to place the photo in the box with a happy face on it if they liked the peer a lot; the neutral face if they sometimes liked the peer; and the sad face if they did not like the peer. A short procedural training session, using photos of play equipment, preceded this task. The third phase consisted of a series of theory of mind tasks drawn from previous research in this area (Dockett 1994). Tasks covered the areas of cognitive connections (Flavell, 1988); appearance-reality distinctions (Flavell, Green & Flavell, 1990) false belief (Perner, 1991) and representational change (Astington & Gopnik 1988). In addition to the cognitive connections taskÐwhich was included to assess children's understanding of perceptual accessÐappearance-reality tasks which used toys and pictures were also included. Previous research (Dockett, 1994; Flavell et al 1990; Wellman, 1990) has indicated that these tasks usually are completed successfully well before an understanding of appearance-reality involving illusory materials. Illusory materials used in appearance-reality tasks included a sponge which appeared to be a rock and a candle which appeared to be a cake. An overview of tasks and questions is included in the Appendix. All interviews with children were conducted in a quiet area of the preschool, which was within sight of the other staff and children. Children were invited to participate in the interviews, which were conducted over a period of several days. Each theory of mind interview lasted about ten minutes and each sociometric rating procedure lasted about fifteen to twenty minutes, including the training session. Interviews followed a standard pattern, however there is also the potential for the researcher to follow the interest of the child by responding to their questions or statements within the interview. Interview sessions were audio taped and then transcribed. Measures A series of measures relating to peer ratings of individual children as well as individual children's responses to the theory of mind tasks were calculated from children's responses in the interviews. In relation to peer popularity, the following measures were generated: a) the number of positive, neutral and negative nominations for each child b) a 'likability score', calculated by subtracting the number of negative ratings from the number of positive ratings and dividing this by the total number of responses (Denham et al. 1990) c) a global or general rating, derived from the likability score (high, neutral or low). Measures calculated in relation to theory of mind tasks were: a)responses by task (cognitive connections; appearance-reality Ðpicture, toy, pretend and illusory object; false beliefÐown and others; representational change) b)an overall score (the sum of correct responses to all theory of mind tasks) c)a global rating, based on responses to theory of mind tasks requiring a representational understanding of mind (high or low) The global rating (high or low) was calculated for theory of mind based on the number of correct responses for specific tasks. Children were rated as high in relation to theory of mind tasks when they successfully completed the tasks related to appearance-reality with illusory objects and false belief. Successful performance on these tasks has been identified as requiring at least the beginnings of a representational theory of mind. Such an understanding is not required to successfully complete the cognitive connections, picture, toy and action-based appearance-reality tasks. Each of these measures was included in a correlational analysis, using SPSS-X. Results. Interview responsesÐpeer popularity Responses from children during the interviews indicated that they had no difficulty understanding the task, identifying their peers or rating them. For example, the following explanations were recorded: Dean: I like to play with Cathy... so I'm going to put her in this one (points to box with smiling face on it and posts photo). I don't like to play with Mandy (places photo in box with sad face). Casey: I don't like Blake. (places photo in box with sad face). That's Dale. I like to play with him every day, cause he's funny (places in box with smiling face). Mary: Greg (looks at photo). He's a really naughty boy ... (posts into box with sad face) Terry: Oh! Casey (looks at photo). I like her a lot (posts in box with smiling face). Mmmm. Mary (looks at photo). I love her (posts into box with smiling face). Ella (looks at photo). I don't play with her at all (posts in box with sad face). (Picks up another photo). Oh, I play with her all right. Only a little bit (posts photo into box with neutral face). But I don't play with him at all (photo of Blake). He's naughty. Neil: Just sometimes I like Evan. Not when he's angry. He gets angry sometimes. When I hit him. Interview responsesÐtheory of mind tasks Children's responses followed patterns identified in previous research (Dockett, 1994). The majority of children responded correctly to the questions relating to cognitive connections, indicating that specific information was conveyed by one sensory modality. For example, the majority of children reported that they did not know what was in the box as they could not see it. Responses to the appearance-reality tasks which used illusory materials demonstrated either a clear understanding that one object could be represented in more than one way, or the conviction that the object a) was really what it looked like, or b) looked like what is was. In the examples that follow, Mandy is not convinced that one object could be represented in more than one way, whereas Alex is equally convinced that this is the case. Researcher:When you look at this, does it look like a cake or does it look like a candle? Mandy:A cake. R:Would you like to pick it up? Is it really a candle, or is it really a cake? M:It's really a cake. Researcher: When you look at this, does this look like a sponge, or does it look like a rock? Alex:A sponge. R:Would you like to pick it up? What is it really, it really a rock or really a sponge? A:A sponge (squeezes it). Soft. Like it's a rock. Popular children and theory of mind In general, children who were mostly rated as disliked by their peers were also classified as performing at a low level in relation to theory of mind tasks (Table 1). Children who were consistently rated as liked by their peers generally performed at a high level on theory of mind tasks. Table 1. General comparison of levels of likability and theory of mind task performance. Global theory of mind rating Global rating of likabilityHighLow High 62 Neutral 12 Low112 Total816 Significant correlations were identified between the overall theory of mind measure (sum of correct responses across all tasks) and the number of positive ratings for individual children (r=.4797, p <.05). In addition, a negative correlation was identified between the overall theory of mind measure and the combined number of neutral and negative ratings (r= -.4769, p. < .05). A similar pattern emerged in relation to the global theory of mind rating (high or low), where this was significantly correlated with the positive ratings of individual children by peers (r= .4257, p. < .05) and with the negative ratings of individual children (r= -.5243 p. <.01). A summary of significant correlations between the global theory of mind rating and sociometric measures is detailed in Table 2. Table 2. Significant correlations between global theory of mind measure and sociometric ratings. Global theory of mind measure Sociometric measures(high/low) Positive nominations (like)r = .4257, p. < .05 Negative nominations (dislike)r = -.5243, p. < .01 Negative and neutral nominations r = -.4087, p. < .05 Global rating (high/neutral/low)r = .4574, p. < .05 Age The only measure significantly correlated with age was performance on false belief tasks overall. Such tasks included questions relating to each child's understanding of their own false belief as well as responses relating to the false beliefs of others. In keeping with the results of previous research (Dockett, 1994),children who were older performed at a higher level on false belief tasks (r = .4277, p. < .05). Gender Gender was significantly correlated with the global theory of mind measure (r = -.4730, p. < .05). A comparison of means for like and dislike measures indicated that, in general, boys were both liked less and disliked more than girls. Correlations calculated separately for groups of boys and girls indicated that the global likability of boys (whether they were rather high, neutral or low in terms of likability) was significantly correlated to their global rating on theory of mind tasks as indicated in Table 3. While the results for girls were in the same directions as those reported for boys, the correlations did not reach levels of significance (Table 3). Table 3. Correlations between gender and theory of mind ratings. Discussion While the small sample size and the exploratory nature of the study represent limitations to the generalisability of results, several patterns have emerged from this study. Firstly, children's performance on the theory of mind tasks presented is similar to that reported in previous research (Dockett, 1994; Flavell, 1988; Wellman, 1990) in that tasks requiring an awareness of cognitive connections and pretense-reality, as well as appearance-reality tasks relying on non-illusory objects, seem to present little difficulty to four-year-olds. In contrast, tasks requiring a representational theory of mind, that is, the awareness that the one object may be represented in different ways and that representations change, present greater difficulty for some four-year-olds. Secondly, it is possible for four-year-olds to rate their peers in terms of likability. The children in this study exhibited no difficulty in performing this task. Rather, they performed this task with ease, often supporting their decisions with comments as to why they were rating a child in a particular way. While not the focus of this paper, it is of interest to note that by about the age of four, some children are already perceived as popular by their peers, whereas others are not so. Positive ratings for individuals (out of a possible 23 nominations) ranged from 4 to 16, while negative ratings also ranged from 4 to 16. The impact of one child being actually disliked by a majority of peers and the reasons for such dislike, are areas of potential research interest which could be pursued in greater depth in the Australian context. As individual children rated their peers, a profile of their own interactions with others became clear. The great variation in this was remarkable, with one child (Alice, aged 57 months) positively nominating only one child, while rating all the other children as ones she did not like to play with. The majority of these children also nominated Alice as someone they did not like to play with. This contrasted with Casey (aged 58 months) who positively rated 9 peers, rated 12 as neutral and negatively rated 2 other children. Again, the implications of such patterns of interaction for young children is a potential area of exploration. Finally, there seems to be a connection between whether or not children are liked and their representational theory of mind. This connection is more pronounced for boys than girls, although the reasons for this are far from clear. As indicated in the introduction, this connection is supported by protagonists of the view that social interaction provides the context for the social construction of understandings and/or the testing of individual understandings. Differences in the interactions of boys and girls (Black, 1992; Dockett, 1994) have been noted and reasons for these proposed, however, any explanation to explain the differences observed in this sample would be speculative. The social interactions of peers have been described as instrumental in alerting children to the possibility that alternative views or perspectives exist (Bonica, 1993; Rayna, Ballion, Breaute & Stambak, 1993). According to this position, children's negotiations with others reveal not only an awareness of other's actions and reactions, but also an awareness of what their partner wants, what they might be thinking and what they will respond to: "in order to have their own ideas accepted, they learn to take the other's wishes and intentions into account" (Rayna et al., 1993, p. 112). Other researchers have stressed the importance of social interactions as a means of creating shared meanings or understandings and communicating about those. GšncŸ (1993) reports that the development of intersubjectivity, defined as "shared understandings of the participants of an activity" (p. 186), promotes joint decision-making, joint construction of meanings and communication about these. Whether or not the role of social interaction is to provide a context in which different perspectives and understandings may be tested, or where shared understandings are constructed, it is the interactions with others that is important. In social interactions, children are likely to act in accord with their own theories of mind and to predict and explain the actions of themselves and others in accordance with the same theories. It is impossible to predict, at this stage, whether it is the representational theory of mind which promotes children's social interactions, or whether such interactions are the basis for developing such theories of mind. References Asher, S. R., Singleton, L. C., Tinsley, B. R. & Hymel, S. (1979). A reliable sociometric measure for preschool children. Developmental Psychology, 15, 443-444. Astington, J. W. & Gopnik, A. (1988). Knowing you've changed your mind: Children's understanding of representational change. In J. W. Astington, P. L. Harris & D. R. Olson, (Eds), Developing theories of mind (pp. 193-206). Cambridge: Cambridge University Press. Astington, J. W. (1991). Intention in the child's theory of mind. In D. Frye & C. Moore (Eds), Children's theories of mind: Mental states and social understanding (pp. 157-172). Hillsdale, NJ: Erlbaum. Black, B. (1992). Negotiating social pretend play : Communication differences related to social status and sex. Merrill-Palmer Quarterly, 38 (2), 212-232. Bonica, L. (1993). Negotiations among children and pretend play. In M. Stambak & H. Sinclair (Eds), Pretend play among 3-year-olds (pp. 55-77). Hillsdale, NJ: Erlbaum. Denham, S. A., McKinley, M., Couchoud, E. A. & Holt, R. (1990). emotional and behavioural predictors of preschool peer ratings. Child Development, 61, 1145-1152. Dockett, S. (1994). Pretend play and young children's developing theories of mind. Unpublished PhD thesis, University of Sydney. Dunn, J. (1993). Young children's close relationships: Beyond attachment. Newbury Park, CA: Sage. Flavell, J. H. (1988). Cognitive connections to mental representations. In J. W. Astington, P. L. Harris & D. R. Olson, (Eds), Developing theories of mind (pp. 244-270). Cambridge: Cambridge University Press. Flavell, J., Green, F., & Flavell, E. (1990). Developmental change in young children's knowledge about the mind. Cognitive Development, 5, 1-27. GšncŸ, A. (1993). Development of intersubjectivity in social pretend play. Human Development, 36, 185-198. Hartup, W. (1989). Behavioural manifestations of children's friendships. In T. J. Berndt & G. W. Ladd (Eds), Peer relationships in child development (pp. 46-70). New York: John Wiley. Howes, C. (1988). Peer interaction of young children. Monographs for the Society of Research in Child Development, 53 (1,serial no 217). Katz, L. (1988). What should young children be learning? Wingspread Journal, (special edition), 200-206. Kemple, K. M. (1991). Preschool children's peer acceptance and social interaction. Young Children, 46 (5), 47-54. Perner, J. (1991). Understanding the representational mind. Cambridge MA: MIT Press. Rayna, S., Ballion, M., Breaute, M. & Stambak, M. (1993). Psychosocial knowledge as reflected in puppet shos improvised by pairs of children. In M. Stambak & H. Sinclair (Eds), Pretend play among 3-year-olds (pp. 79-113). Hillsdale, NJ: Erlbaum. Rogoff, B. (1990). Apprenticeship in thinking: Cognitive development in social context. New York: Oxford University Press. Wellman H. M. (1990). The child's theory of mind. Cambridge, MA: MIT Press. Appendix Cognitive connections Can you see this box/? Tracey has put something inside it, and I didn't see her do it, so I don't know what it is. 1. Do you know what it is? 2. Do I know what is inside the box? 3. Let's shake it. Can you hear it? 4. Can I hear it? 5. Can you see it? 6. How can we find out what is inside the box/tin? Appearance-reality tasks Appearance question 1. When you look at this, does it look like a rock or does it look like a sponge? Reality question 2. Is it really a sponge, or is it really a rock? Picture tasks Appearance question 1. When you look at this, does it look like shoes? Reality question 2. For real, is this really shoes, or really a picture? Toy tasks Appearance question 1. When you look at this, does it look like a car? [children invited to pick up item] Reality question 2. For real, is this really a car or really a toy? Pretend tasks Appearance question 1. When you look at me, does it look like I'm brushing my teeth? Reality question 2. For real, am I brushing my teeth, or pretending to brush my teeth? Do you remember this? This is really a candle but it looks like a cake. [the researcher puts the candle in front of her mouth, as if she were eating it] Appearance question 1. When you look at me with your eyes, does it look like I'm eating this? Reality question 2. For real, and I pretending to eat this, or am I really eating it? False belief and representational change tasks Appearance question 1. Look at this. Does it look like a packet of biscuits, or does it look like a packet of noodles? Let's open it up and have a look. Reality question 2. What is it really? Is it really a packet of noodles or is it really a packet of biscuits? False belief question 3. X (name of child's friend who has not yet been interviewed) hasn't seen this box. If we close this box up again (contents replaced in packet, flap closed), and ask X to look at it, what will X think is inside it? Will s/he think it is a packet of biscuits or will s/he think it is a packet of noodles? Representational change question 4. When you saw the packet, what did you think was inside it before I opened it up