Journal of Experimental Child Psychology 126 (2014) 52–67

Contents lists available at ScienceDirect

Journal of Experimental Child Psychology journal homepage: www.elsevier.com/locate/jecp

Promoting theory of mind during middle childhood: A training program Serena Lecce a,⇑, Federica Bianco a, Rory T. Devine b, Claire Hughes b, Robin Banerjee c a

Department of Brain and Behavioral Sciences, University of Pavia, 27100 Pavia, Italy Department of Psychology, University of Cambridge, Cambridge CB2 3EB, UK c School of Psychology, University of Sussex, Brighton BN1 9QH, UK b

a r t i c l e

i n f o

Article history: Received 21 November 2012 Revised 6 March 2014

Keywords: Theory of mind Intervention Middle childhood Maintenance Executive functions Conversational approach

a b s t r a c t Evidence that conversations about the mind foster improvements in theory of mind (ToM) is growing, but their efficacy in typically developing school-aged children has yet to be demonstrated. To address this gap, we designed a conversation-based training program for 9- and 10-year-olds and measured its effectiveness by pre- and post-test comparisons of performance on age-appropriate ToM tasks for two groups (matched at pre-test for gender, age, socioeconomic background, verbal ability, reading comprehension, executive functions, and ToM) who were assigned to either the intervention condition (n = 45) or an active control condition (n = 46). The intervention group showed significantly greater gains in ToM than the control group; this contrast was stable over 2 months, and (in a subsample) the improvement in ToM was independent of any changes in executive functions. Implications for the role of conversations about the mind in children’s mental state reasoning are discussed. Ó 2014 Elsevier Inc. All rights reserved.

Introduction Children’s theory of mind (ToM) represents one of the liveliest areas of study in developmental psychology. After 30 years of research in this field, there is some consensus about key milestones of ⇑ Corresponding author. Fax: +39 0382986132. E-mail address: [email protected] (S. Lecce). http://dx.doi.org/10.1016/j.jecp.2014.03.002 0022-0965/Ó 2014 Elsevier Inc. All rights reserved.

S. Lecce et al. / Journal of Experimental Child Psychology 126 (2014) 52–67

53

normative ToM development, particularly during the preschool years (Wellman, Cross, & Watson, 2001). Recent years have also seen mounting evidence that ToM continues to develop during and even beyond the school years (Miller, 2012), extending into adolescence and adulthood (Apperly, 2011; Devine & Hughes, 2013). During the school years, children become increasingly sophisticated in applying their ToM skills to make sense of complex social situations. Despite being able to pass second-order false belief tasks (involving mistaken beliefs about beliefs), 9-year-olds are still only beginning to effectively use their ToM skills to explain behavior (Banerjee, Watling, & Caputi, 2011). Recent studies have shown that, with increasing age, children became better at reasoning about beliefs and perspectives and in applying these skills in a variety of contexts/scenarios. For example, they become able to accurately judge one person’s beliefs about the intentions of others (Miller, 2009; Pillow, 1991) and use a higher frequency of mental state terms to describe social behavior (Meins, Fernyhough, Johnson, & Lidstone, 2006). Between 9 and 11 years of age, children become more sophisticated in interpreting ironic utterances (Filippova & Astington, 2008), and understanding faux pas (Baron-Cohen, O’Riordan, Stone, Jones, & Plaisted, 1999). Crucially, a recent study of 230 children between 8 and 13 years of age showed age-related improvement in two ToM tasks: the well-established text-based Strange Stories task (White, Hill, Happé, & Frith, 2009) and a novel paradigm involving brief clips from a classic silent film (Devine & Hughes, 2013). This result is important because it suggests that children’s ability to use their understanding of mental states continues to improve beyond the preschool years. Despite these promising studies, we still know very little about what drives development in ToM beyond early childhood. The current article addresses the hypothesis that conversations about mental states have a crucial role in the development of relatively advanced ToM skills. Conversational approach to ToM development Several strands of evidence, including cross-cultural studies (Hughes et al., 2014; Lecce & Hughes, 2010) and observational studies (see Hughes, 2011), indicate that variation in social experiences contributes to individual differences in ToM. Interestingly, empirical results from twin studies suggest that as children increase in age, the social environment becomes important in explaining individual differences in ToM (Hughes & Cutting, 1999; Hughes et al., 2005; Ronald, Viding, Happé, & Plomin, 2006). Notably, a large-scale study of 5-year-old twins showed common influences of shared environment on individual differences in children’s understanding of false beliefs and verbal ability, indicating that variation in linguistic environments (i.e., family talk) may contribute to variation in children’s ToM (Hughes et al., 2005). Numerous authors have highlighted participation in conversations about mental states as an important influence on children’s development of ToM. According to this conversational approach, exposure to conversations that are rich in reference to (and explanation of) mental states such as desires, emotions, and beliefs facilitates children’s understanding of others’ minds (Dunn & Brophy, 2005; Nelson, 2005; Turnbull & Carpendale, 1999). Indeed, some existing theoretical models present mental state conversations as the key learning context within which ToM progress can be made. For example, Nelson (2007) proposed that conversations make children enter into the ‘‘community of minds,’’ allowing them to reflect on their social experiences and improve their awareness that people can have different mental states that relate to the same situation. From this perspective, conversations constitute a privileged context for helping children to reflect on the differences between others’ and their own states of mind. In turn, the need to coordinate others’ points of view with their own experience gives birth to a gradually constructed understanding of the mind (Harris, 1999). Direct evidence for the role of mental state talk in school-aged children’s ToM skills comes from three separate sources. First, extending early demonstrations of delayed ToM success among deaf children born to hearing parents (DoH) but not deaf children born to deaf parents (DoD) (e.g., Peterson & Siegal, 1995; Peterson & Slaughter, 2006), two studies showed that the richness of conversational experience at school predicts variation in ToM success within DoD children (Meristo et al., 2007; Tomasuolo, Valeri, Di Renzo, Pasqualetti, & Volterra, 2013). Second, longitudinal studies of typically developing children have shown that early variation in mothers’

54

S. Lecce et al. / Journal of Experimental Child Psychology 126 (2014) 52–67

mental state talk predict individual differences in children’s ToM both in preschool (Ensor & Hughes, 2008; Ruffman, Slade, & Crowe, 2002) and during middle childhood (Ensor, Devine, Marks, & Hughes, 2014). Finally, recent training studies with young children suggest a causal influence of children’s participation in mental state conversations on ToM development. In two separate studies, Ornaghi and colleagues reported that preschoolers and Year 1 children who had opportunities to take part in conversations about mental states showed better ToM than children who did not (Ornaghi, Brockmeier, & Grazzani, 2011, 2014). Within the same conversational paradigm, Ziv, Smadja, and Aram (2013) showed that promoting sociocognitive themes within conversations between parents and children during shared reading led to enhancements in children’s mental state understanding. Together, these studies suggest that young children’s participation in conversations about mental states brings about fundamental changes in their capacity to reason in psychological terms (de Rosnay & Hughes, 2006). What is lacking, however, is direct experimental evidence for a similar influence of mental state conversations on ToM development during middle childhood, which is known to be significantly associated with real-world indicators of social success (e.g., Banerjee et al., 2011; Caputi, Lecce, Pagnin, & Banerjee, 2012). Addressing this gap was a key goal for the current study. Specifically, we designed an intervention to demonstrate specific causal relations between mental states conversations and the development of ToM understanding in 9- and 10-year-olds. Before entering into the details of the training program literature, it should be stressed that middle childhood is also characterized by growth in executive functions (see Davidson, Amso, Anderson, & Diamond, 2006; Huizinga, Dolan, & van der Molen, 2006), which are known to be crucial for ToM abilities (Carlson & Moses, 2001; Hughes, 1998b). Therefore, it is possible that executive function skills might also underpin the continued growth in ToM use during the school years. This acknowledgment of a possible link between ToM and executive functions during middle childhood raises two important issues in relation to any ToM training study. First, given that executive functions promote learning across a number of domains (e.g., Blair & Razza, 2007; Bull & Scerif, 2001; Espy et al., 2004), individual differences in executive functions prior to training might predict the extent to which children’s ToM improves during the training. By this account, we should expect significant associations between individual differences in executive functioning at pre-test and children’s gains in ToM as a result of the training. Alternatively, it might be that ToM training is effective not because it directly causes improvement in ToM but rather because it enhances children’s executive functions. If this is the case, we should find significant associations between gains in executive functions and gains in ToM scores. The current study was designed to help us address these possibilities as well as the more fundamental hypothesis that a conversation-based training program would improve ToM during middle childhood. ToM training programs Existing studies with typically developing children have shown that training ToM can be successful during the preschool years (Kloo & Perner, 2008). In two early studies on this topic, Slaughter and Gopnik (1996) and Appleton and Reddy (1996) showed that 3- and 4-year-olds involved in conversations about beliefs, desires, and perceptions improved their false belief understanding more than children in the control group. Two subsequent studies by Hale and Tager-Flusberg (2003) and Lohmann and Tomasello (2003) highlighted the importance of using sentential complement constructions (i.e., sentences that take a full clause as their object complement, e.g., ‘‘Peter thinks that the candy is in the cupboard’’) to improve 3- to 5-year-olds’ false belief understanding. Importantly, other studies have confirmed the importance of giving detailed feedback and explanations to children on why their answers to ToM questions were right or wrong during verbal interactions (Clements, Rustin, & McCallum, 2000; Melot & Angeard, 2003). Although our own efforts focused on using conversations to stimulate progress in more advanced aspects of ToM reasoning, the studies cited above provide important insights into the principles of effective ToM training. Therefore, in designing our intervention, we focused on mental state terms within sentential complement constructions and made frequent use of feedback

S. Lecce et al. / Journal of Experimental Child Psychology 126 (2014) 52–67

55

and explanation within group conversation. We chose to use group conversations about mental states because there is strong agreement in the literature that the frequency, quality, and content of conversations about the mind predict later ToM skills (Appleton & Reddy, 1996; Ensor & Hughes, 2008; Ornaghi et al., 2011; Peterson & Slaughter, 2003). Notably, the emphasis in the conversations was on elaborating on children’s comments, explaining the reasons why their answers were right or wrong, and highlighting the existence of different points of view on the same event. We designed two types of activity to prompt conversations about mental states: one centered on short stories/vignettes and the other focused on the meaning of mental state verbs. Although the comprehension of mental state verbs is a key correlate of ToM (Grazzani & Ornaghi, 2012), stories offer a good opportunity to reflect, as a ‘‘social apprentice,’’ on the range of complex mental state dynamics relevant to everyday social contexts (Dyer, Shatz, & Wellman, 2000; Lecce, Zocchi, Pagnin, Palladino, & Taumoepeau, 2010). Our training program also built on research regarding key developmental changes in ToM skills during middle childhood and highlighted the dynamic nature of mental states. Being able to recognize that mental states are not static but rather change over time is one of the core acquisitions within ToM (Flavell, 1993; Gopnik & Astington, 1988). Given that the maintenance of a training effect is crucial in the evaluation of the intervention itself, it is surprising that this issue has rarely been addressed in previous studies. Indeed, most published research studies have measured children’s ToM skills within 1 week of the intervention (Clements et al., 2000; Hale & Tager-Flusberg, 2003; Lohmann & Tomasello, 2003; Melot & Angeard, 2003; Slaughter & Gopnik, 1996), precluding any examination of whether any detected effects endured beyond training. In one notable exception, Appleton and Reddy (1996) completed both an immediate post-test and a follow-up session 2 to 3 weeks after the training program, showing that at follow-up children in the training group outperformed those in the control group on all of the tasks. This finding indicates that the effects of conversation-based intervention may remain stable for some weeks after the end of the intervention. However, even 3 weeks represents a relatively short period of time to establish the efficacy of an intervention. Thus, as well as testing children’s ToM 2 weeks after the end of the training in our studies, we also conducted a 2-month follow-up to test the enduring effects of the intervention program. The current study This study had three main aims. The first goal was to test the efficacy of our ToM intervention and provide experimental evidence for the role of conversations about mental states in the development of ToM skills during middle childhood. The second goal was to test the stability of the training effect by retesting children’s ToM 2 months after the end of the training program. The third goal was to control for possible variations in executive functions when evaluating the impact of the training conversations about mental states. In addressing these issues, we also controlled for individual differences on a wide variety of potential confounding variables that are known to be associated with ToM: children’s family background (Cutting & Dunn, 1999), verbal ability (Milligan, Astington, & Dack, 2007), and reading comprehension (Lecce & Hughes, 2014). Method Participants We recruited a sample of 91 children from two primary schools in the area surrounding Milan in northern Italy. Participants had a mean age of 9 years 7 months (SD = 4.03 months) and were randomly assigned to either the ToM condition (30 boys and 15 girls) or the control condition (24 boys and 22 girls). A total of seven classes took part in the current study. Children within the same classroom were assigned to different conditions. All of the children were fluent Italian speakers. Criteria for inclusion were written parental consent and absence of cognitive or learning difficulties.

56

S. Lecce et al. / Journal of Experimental Child Psychology 126 (2014) 52–67

A subsample of 46 children (Mage = 9 years 7 months, SD = 3.8 months) were evaluated on executive functions not only at pre-test but also at post-test and follow-up. Of these children, 21 (15 boys and 6 girls) were in the ToM condition and 25 (13 boys and 12 girls) were in the control condition. Measures Pre-test Socioeconomic status. We administered the Italian translation of the Family Affluence Scale (Torsheim, Boyce, Currie, & Zambon, 2006). The Family Affluence Scale is a four-item measure of family wealth. The children completed a short questionnaire about family car ownership (range = 0–2), having their own unshared room (range = 0–1), the number of computers at home (range = 0–3), and the number of times they went on a holiday during the past year (range = 0–3). Children’s scores were aggregated into an overall index of family socioeconomic status (range = 0–9). Verbal ability. We used the Vocabulary subtest of the Wechsler Intelligence Scale for Children–Revised (WISC-R; Italian version: Orsini, 1997). We asked children to define a series of 32 words. We scored responses using the standard 3-point scale: 0 for wrong or extremely weak answers, 1 for correct but poor answers, and 2 for fully correct definitions. Scores could range from 0 to 64. Reading comprehension. We measured children’s ability to understand the meaning of a written text through the Memory and Transfer Standardized Reading Comprehension battery (Cornoldi & Colpo, 1998). Children were required to read a passage silently and answer 10 multiple-choice questions requiring inferential reasoning. Participants were allowed to consult the text during the answering phase, and no time limit was imposed. Scores could range from 0 to 10. Pre-test, post-test, and follow-up Executive functions. We tested children’s executive functions using a modified version of the Tower of London task (Shallice, 1982) to index planning and the Backward Digit Span task from the Italian version of the WISC-R (Orsini, 1997) to measure working memory. In our modified version of the Tower of London task, we showed children a series of images depicting the starting and final configurations of three colored balls on three pegs of different heights. Children were told that the big peg could carry all three balls, the middle peg could carry two balls, and the little peg could carry just one ball and that only one ball could be moved at a time. Children were asked to work out and write down the total number of moves needed to reach the final configuration for each trial. Children were given problems with three, four, or five expected moves. Each trial was scored as pass (1) or fail (0), and total scores could range from 0 to 7. In the Backward Digit Span task, children listened to a series of digit sequences and were asked to recall them in reverse order. Children completed seven sequences of digits. The number of digits in each sequence increased from two to eight. Each sequence was marked as a pass (1) or fail (0), and total scores could range from 0 to 7. Theory of mind. We tested children’s ToM at pre-test, post-test, and follow-up using parallel versions of the Strange Stories task (Happé, 1994; White et al., 2009). We administered an Italian translation of the Strange Stories task that has been used in other studies (Cavallini, Lecce, Bottiroli, Palladino, & Pagnin, 2013; Lecce et al., 2010). The Strange Stories task is an advanced test of ToM that assesses the ability to make inferences about mental states by interpreting nonliteral statements. We selected a subgroup of six mentalistic stories involving double bluffs, misunderstandings, white lies, and persuasion. After reading the stories, we asked children to explain a character’s behavior. Children recorded their answers in writing without a time limit. In line with scoring guidelines (White et al., 2009), we rated responses using a 3-point scale: 0 for incorrect answers, 1 for partially correct answers, and 2 for full and explicit answers. An example of the scoring criteria is given in Appendix A. Two raters independently coded 25% of the responses at

S. Lecce et al. / Journal of Experimental Child Psychology 126 (2014) 52–67

57

each time point, and interrater agreement was established using Cohen’s kappa (at pre-test, j = .83; at post-test, j = .76; at follow-up, j = .84). All remaining responses were coded jointly by the raters. Disagreements were resolved through discussion between the raters. Total scores ranged from 0 to 12. Design and procedure Children were pre-tested individually at school on ToM tasks and on a series of control variables: verbal ability, reading comprehension, planning, and working memory. Information about family background was also gathered during pre-test. After the pre-test, children took part in a training program that consisted of four sessions. At the end of the training program, all children were post-tested twice to investigate the stability of training effects on ToM. The first post-test session took place between 4 and 19 days after the end of the training program (M = 11.42 days, SD = 6.0). The second post-test session took place between 55 and 63 days after the end of the training program (M = 59.21 days, SD = 1.5). As noted earlier, within each participating class, there were children belonging to the two training conditions. So, although children were instructed not to talk about the training after the intervention, cross-contamination between the groups was still possible. This, however, does not represent a major concern in terms of results because if there was crosscontamination it would reduce (rather than increase) the effects of training. Training procedure The structure of the ToM and control training programs consisted of four sessions, each involving group conversations about two stories and two language exercises. Thus, children received eight stories (two stories in each of the four training sessions) and eight language exercises (two exercises in each of the four training sessions) in total. Sessions lasted from 40 to 50 min and took place in a separate room located in children’s primary school. In each trial, a researcher presented children with a written vignette and then asked them to answer four or five questions concerning the story. The researcher asked children to write down their answers to the questions individually and then encouraged them to take part in a group conversation. During this conversation, the researcher used the story questions as prompts and ensured that all children took part in the conversation, discussing their points of view on the story. The researcher also made frequent use of positive and corrective feedback, expanding children’s comments and explaining the reasons why their answers were right or wrong. When all of the answers had been discussed, the researcher made a final comment highlighting the core characteristics of the story. The researcher then moved toward the second part of the trial, which consisted of a language exercise. Children were individually presented with a sentence taken from the story and were encouraged to find a synonym of a given verb in the sentence by selecting one of four given alternatives. When all children had written down their own answers, the experimenter opened a group conversation on the meaning of the chosen verb based on individual responses. Again, the experimenter made extensive use of feedback during this group conversation. Below we give more details on the content of the training activities in the two conditions. In the experimental condition (ToM condition), we developed stories based on the mental state stories of the revised Strange Stories task (White et al., 2009). More specifically, we developed two misunderstanding, two irony, two faux pas, and two double-bluff stories. We selected these stories because in each story the main characters have different points of view on the core event of the narrative. This discrepancy in beliefs/knowledge created a problem between the main characters and, thus, represented the focus of the training activities. The stories we devised were similar to the Strange Stories in content, difficulty, and length, but the intervention program was distinctive in terms of the number and type of questions children were asked to answer. In particular, the Strange Stories test requires children to explain why the main character said

58

S. Lecce et al. / Journal of Experimental Child Psychology 126 (2014) 52–67

something that is not literally true, leaving most of the aspects of the social scenarios implicit and undisclosed. In contrast, in the intervention stories, we asked children a series of questions tapping different dimensions of the situations to elicit a complete and explicit understanding of the mental states and behavior of characters. Questions were asked about (a) the main character’s mental state, (b) one character’s belief about the other character’s mental state, and (c) mental states underlying social behavior. We designed an additional question for each story to highlight the dynamic nature of mental states. We asked children what the main character could do or say to change the other character’s mental state and, thus, reduce the difference in perspective. A researcher led the group conversation regarding each of these questions and closed the conversation with a statement that stressed the core dimension of each type of story (misunderstanding, irony, faux pas, and doublebluff) and highlighted the difference in perspective between participants. We then asked children to imagine an episode similar to the one presented in the story and to consider how other people’s points of view could be modified (e.g., to correct the misunderstanding) in that situation. We created the language exercises in the ToM condition based on the metacognitive language task by Olson and colleagues (Olson, Antonietti, Liverta Sempio, & Marchetti, 2006). To evaluate knowledge of mental state verbs, children were asked to choose a synonym for the verbs think and say by selecting one of four options. In the training program, we selected a sentence from the story that contained one of the following mental state verbs: think, believe, and imagine. We then asked children to find the most appropriate synonym for these verbs, based on the general meaning of the story, by selecting one of four given options. In each session, the experimenter made extensive use of mental state verbs (e.g., think, know) within sentential complement construction. Appendix B contains a description of one complete trial. We developed stories for the control condition similar to those used in the control vignettes of the Strange Stories task (White et al., 2009). These stories described physical events and involved human characters. As in the ToM condition, each story was followed by a series of questions. The first three questions were about specific facts of the story or details presented in the text. The last question of each story was about a physical event not explicitly mentioned in the text. Thus, to give the correct response, children needed to make a physical inference and go beyond explicit information given in the text. Each of these questions was discussed in groups led by the researcher. At the end of the conversation, the researcher made a final comment starting from the last question and stressing the importance of the inference in order to understand the story. Children were then asked to recall an episode similar to the one presented in the story and to tell it. An example trial is described in Appendix B. The language exercise in the control condition had a structure parallel to that used for the mental verbs in the ToM condition and involved the following physical verbs: receive, give, take, let, do, and call. Overall, the intervention focused on going beyond the explicit information in a text to make links between different parts of a text. During group conversations the researcher made no use of mental state verbs. Results Preliminary analysis showed no effects of gender. Thus, given that we had no a priori hypothesis, gender was excluded from the main analysis reported below. Table 1 shows descriptive statistics for all variables in the analysis. t-Tests showed no significant differences between the experimental and control groups on any of the pre-test control measures, all ps > .17. Importantly, the two groups also had a similar level of ToM ability at pre-test, t(89) = 0.33, p = .744, 95% confidence interval (CI) [ 0.99, 0.71]. To investigate the efficacy of our training, we conducted a mixed analysis of variance (ANOVA) on the Strange Stories scores with time (pre-test, post-test, or follow-up score) as the within-participants factor and training group (ToM or control) as the between-participants factor. There were significant main effects of time, F(2, 178) = 129.74, p < .001, partial g2 = .59, and training group, F(1, 89) = 11.44, p = .001, partial g2 = .11. Crucially, there was also a significant time by group interaction,

59

S. Lecce et al. / Journal of Experimental Child Psychology 126 (2014) 52–67 Table 1 Descriptive statistics and group differences on control and ToM measures. ToM group (n = 45)

Pre-test age in months Socioeconomic background (0–9) Verbal ability (0–64) Reading comprehension (0–10) Pre-test planning (0–7) Pre-test working memory (0–7) Pre-test ToM–Strange Stories (0–12) Post-test age in months Post-test ToM–Strange Stories (0–12) Follow-up age in months Follow-up ToM: Strange Stories (0–12) Gains in ToM–Strange Stories from pretest to post-test Gains in ToM–Strange Stories from pretest to follow-up

Control group (n = 46)

M (SD)

Median

Actual range

M (SD)

Median

Actual range

115.22 (3.7) 5.8 (1.8) 38.44 (9.3) 6.20 (2.4) 4.78 (2.4) 3.27 (1.4) 6.56 (1.8) 116.56 (3.9) 9.98 (1.5) 118.12 (3.8) 10.40 (1.4) 3.42 (1.7)

115 6 40 7 6 3 7 116 10 118 11 3

109–122 1–9 17–58 1–10 0–7 1–7 3–10 111–123 6–12 112–125 6–12 0–7

116.17 (4.3) 5.70 (1.6) 38.41 (7.1) 6.52 (2.4) 4.04 (2.5) 3.26 (1.4) 6.70 (2.2) 117.30 (3.6) 8.22 (2.0)⁄⁄⁄ 118.87 (3.7) 8.98 (1.8)⁄⁄⁄ 1.52 (2.0)⁄⁄⁄

116 6 39 7 5 3 6 117 8 119 9 1

109–130 3–9 21–56 2–10 0–7 1–6 3–11 111–123 4–12 112–124 5–12 2–6

0–8

2.28 (2.3)⁄⁄

3.84 (2.0)

4

Note. Asterisks indicate significant t tests of ToM versus control group comparisons:

⁄⁄

p < .01;

⁄⁄⁄

3

4–7

p < .001.

F(2, 178) = 12.62, p < .001, partial g2 = .12. To break down this interaction, we performed pairwise contrasts applying a Bonferroni correction for multiple comparisons. These contrasts revealed significant differences between the ToM and control groups at post-test, p < .001, 95% CI [1.04, 2.48], and followup, p < .001, 95% CI [0.74, 2.10]. Moreover, as shown in Table 1, the improvements between pre-test and post-test and between pre-test and follow-up in the ToM group were significantly greater than the corresponding improvements in the control group, t(89) = 4.85, p < .001, d = 1.02, 95% CI [1.12, 2.68], and, t(89) = 3.44, p = .001, d = 0.72, 95% CI [0.66, 2.46], respectively.1 It should be noted that this pattern of results cannot be attributed to a small number of outliers: when the 11 children in the experimental group who had the highest gain scores from pre-test to post-test (gains of 5, 6, or 7) were excluded (i.e., removing the top 24% of the group), we still found a significantly greater gain from pretest to post-test in the ToM group than in the control group, t(78) = 2.98, p = .004, d = 0.70, 95% CI [0.38, 1.93]. To investigate whether our pre-test control variables (socioeconomic background, age, verbal ability, reading comprehension, planning, and working memory) moderated the effect of the ToM training, we conducted two separate hierarchical moderated regressions on ToM gain scores from pre-test to post-test sessions and from pre-test to follow-up. At Step 1 we entered the training condition and the pre-test variables (mean centered), and at Step 2 we entered the interaction effects between pre-test variables and training condition. Results showed that Step 1 was significant in both regressions, F(7, 83) = 3.51, p = .002 (for pre-test vs. post-test gain score), and F(7, 83) = 2.81, p = .011 (for pre-test vs. follow-up gain score), with training condition being the only significant predictor, b = .47, p < .001 (for pre-test vs. post-test gain score), and b = .38, p < .001 (for pre-test vs. follow-up gain score). Furthermore, Step 2 did not lead to a significant increase in variance explained, DF(6, 77) = 0.44, p = .853 (for pre-test vs. post-test gain score), and DF(6, 77) = 0.78, p = .589 (for pre-test vs. follow-up gain score). Thus, training effects were not moderated by any of the covariates. Next, we evaluated the extent to which the pattern of changes in Strange Stories was mirrored by the changes on executive function tasks of planning and working memory. Table 2 shows descriptive statistics for the subsample of children who completed the executive function tasks at pre-test,

1

Note that gains from post-test to follow-up were not significantly different between groups, t(89) = 0.94, p = .348.

60

S. Lecce et al. / Journal of Experimental Child Psychology 126 (2014) 52–67

Table 2 Descriptive statistics and group differences on executive function scores at pre-test, post-test, and follow-up. ToM group (n = 21)

Pre-test planning (0–7) Pre-test working memory (0–7) Post-test planning (0–7) Post-test working memory (0–7) Follow-up planning (0–7) Follow-up working memory (0–7) Gains in planning from pre-test to post-test Gains in planning from pre-test to follow-up Gains in working memory from pretest to post-test Gains in working memory from pre-test to follow-up

Control group (n = 25)

M (SD)

Median

Actual range

M (SD)

Median

Actual range

3.10 3.67 4.76 5.57 5.29 5.05 1.67

(2.5) (1.2) (2.0) (1.6) (1.6) (1.5) (2.9)

2 3 5 6 6 4 1

0–7 2–6 1–7 2–7 2–7 3–7 3–7

2.64 3.52 3.96 4.52 3.52 4.88 1.32

2 3 4 5 4 5 1

0–7 1–6 0–7 0–7 0–6 1–7 1–6

2.19 (2.4)

2

2–7

0.88 (2.2)+

1

4–5

1.90 (1.3)

2

1.00 (1.8)+

1

3–5

1.38 (1.6)

1

1.36 (1.9)

1

1–5

0–4 2–4

(2.5) (1.4) (2.2) (2.2)+ (2.02)⁄⁄ (2.0) (1.8)

Note. Asterisks indicate significant t tests of ToM versus control group comparisons: +p 6 .10;

⁄⁄

p < .01.

post-test, and follow-up. Specifically, we conducted two separate mixed ANOVAs on the Tower of London planning and on WISC working memory scores with time (pre-test, post-test, or follow-up score) as the within-participants factor and training group (ToM or control) as the between-participants factor. On planning, results revealed a significant main effect of time, F(2, 88) = 13.26, p < .001, partial g2 = .23, but neither the main effect of group, F(1, 44) = 3.87, p = .055, nor the interaction effect, F(2, 88) = 2.00, p = .142, was significant. On working memory, analyses showed a significant main effect of time, F(2, 88) = 20.29, p < .001, partial g2 = .32, but again the interaction effect, F(2, 88) = 2.03, p = .137, and the main effect of group, F(1, 44) = 1.26, p = .268, were not significant. As a further direct test to rule out the possibility of gains in executive functions being responsible for the training effects on ToM, we conducted an analysis of covariance (ANCOVA) on the gains in Strange Stories scores from pre-test to post-test and from pre-test to follow-up. In each ANCOVA, training group (ToM or control) was the between-participants factor, and the corresponding gains in both planning and working memory were included as covariates. In both cases, there remained a significant effect of training group on the gains in Strange Stories, F(1, 42) = 8.71, p = .005, partial g2 = .17, and F(1, 42) = 5.95, p = .019, partial g2 = .12. Indeed, further analysis confirmed that the gains in Strange Stories scores were not correlated with the gains in executive function scores.

Discussion This study evaluated the effects of a ToM training program based on conversations about mental states in typically developing school-aged children. Compared with the control group (matched for gender, family socioeconomic status, age, verbal ability, reading comprehension, planning, and working memory and pre-test ToM performance), children in the experimental group performed significantly better on the ToM task at post-test and follow-up and made greater gains in ToM from pre-test to post-test and from pre-test to follow-up. Crucially, these results were independent of individual differences on all pre-test measures (age, planning, working memory, reading comprehension, socioeconomic status, and verbal ability) and remained significant once the effects of improvements in planning and working memory were taken into account. The sustained group difference at the 2-month follow-up indicates that our training program had an enduring impact on ToM performance. Our positive findings are in line with reports that

S. Lecce et al. / Journal of Experimental Child Psychology 126 (2014) 52–67

61

conversation-based interventions are effective for improving preschoolers’ understanding of mind (e.g., Appleton & Reddy, 1996). The use of a very closely matched active control condition strengthens this conclusion; whereas the control group in Appleton and Reddy’s (1996) study did not take part in conversations but rather simply listened to stories, our control group engaged in conversations for the same amount of time as the training group. Consistent with other findings from longitudinal research that highlight a predictive association between early conversations involving mental states and later individual differences in ToM (for a review, see de Rosnay & Hughes, 2006), therefore, we propose that the content of the conversations (rather than their simple presence) underpinned the observed gains in ToM performance. Specifically, we believe that our training program helped children to become more sensitive, motivated, and aware of when and how to use their insights about mental states such as beliefs. It is likely to have strengthened children’s ability to make mental inferences in a context-sensitive manner, to detect relevant information in social situations, and to construct appropriate models of complex social interactions. The evidence that interventions are effective in changing children’s ToM performance not only in preschool, when children are facing the core acquisition of false belief understanding, but also later on in development is an important result. Indeed, it provides crucial support for recent empirical data (Banerjee et al., 2011; Ensor et al., 2014) and emerging theoretical proposals (e.g., Hughes, 2011) that social and communicative experiences continue to provide a kind of social apprenticeship during middle childhood. We have made the first step in testing the existence of causal relations between schoolaged children’s conversations about mental states and ToM through a carefully controlled intervention design. The next step, we believe, is to elucidate the specific social processes that are responsible for improvements in ToM. Specifically, we hope that our work will open the way for a new wave of research into the specific interactive processes that underpin children’s progress in reasoning about mental states. The details of which particular features of social interaction are most important remain opaque at this time, but we do have some intriguing indications from the results obtained so far. For example, Banerjee and colleagues (2011) showed that variations in active rejection by peers, rather than levels of peer acceptance, were a key antecedent of ToM understanding during middle childhood, implying that limited participation in social interactions with peers may be constraining children’s opportunities to learn about mental phenomena. Indeed, our work showed the importance of having the opportunity to reflect and speak about mental phenomena, consistent with the extensive body of evidence from younger children showing that conversations about mental states (e.g., beliefs, feelings, desires) are predictive of enhanced social understanding (Dunn, 1996; Dunn, Brown, & Beardsall, 1991; Hughes, 2011). However, the current study was not designed in such a way that we could pinpoint the specific features of conversations about the mind that mediate the impact of social interactions (whether in a training program or in naturalistic peer interaction) on ToM. A range of factors could be involved, and these factors are not mutually exclusive: exposure to mental state language during interpersonal discourse, increased social motivation to understand others’ behaviors (e.g., to facilitate peer group entry and successful play), and the provision of directive feedback that challenges children’s misconceptions and erroneous assumptions about another person’s mental state. Each of these features of the group conversations in our training programs is conceivably also important in children’s naturalistic discourse with social partners. In addition to determining which aspects of social interactions are most important, future studies should examine moderating variables that may heighten or decrease the influence of social experience on ToM. For example, children might respond differently to social input depending on the nature of the given social partners (e.g., peer vs. adult, level of closeness or familiarity, within-family vs. extra-familial interactions) (Cutting & Dunn, 2006; Dunn, 1996; Hughes, Marks, Ensor, & Lecce, 2010). In addition, both individual differences between children and contextual variations in social circumstances (e.g., socioeconomic status, culture) might constrain or enhance children’s receptiveness to social input regarding mental states (Cutting & Dunn, 1999; Lecce & Hughes, 2010). Moreover, although we did not find a role of executive functions within the training effects in this study, a substantial literature points to the importance of executive skills in ToM (Apperly, 2011; Hughes, 1998a)

62

S. Lecce et al. / Journal of Experimental Child Psychology 126 (2014) 52–67

and further work on the interface among social experience, executive functions, and ToM is clearly needed. Caution is warranted in the interpretation of our findings because more research is required to make the program fully adherent to criteria of evidence-based interventions (Flay et al., 2005). First, the number of children involved in the studies remains fairly small, such that more research is needed to confirm our results. The limited variation within our sample’s socioeconomic status and the restricted range of measures used to control for executive functions also point to the need for a larger study. Further research is also needed to investigate the longer term durability of effects beyond 2 months because this is still a relatively short period of time. Notably, the impact of our training on other ToM measures clearly needs attention in future research, as does the broader range of outcome measures that might follow from progress in ToM. Given recent empirical findings, good candidates for such wider outcomes include children’s school achievement (Lecce, Caputi, & Hughes, 2011; Lecce, Caputi, & Pagnin, 2014), metacognition (Lecce et al., 2010; Lockl & Schneider, 2007), and peer relationships (Banerjee et al., 2011; Caputi et al., 2012). It will be important for future research to incorporate these kinds of measures, together with the exploration of mediator and moderator variables, in order to evaluate whether our training program is effective in promoting children’s more general level of adjustment via improvements in ToM. In addition, it would be interesting in future studies to include a measure of children’s understanding of mental state talk in order to disentangle the effect of conversation from that of the language exercise and investigate what aspect of the training is more relevant in driving ToM changes. In conclusion, the current findings extend the existing literature on ToM interventions for younger children by showing that our ToM training program was effective in stimulating gains in ToM that were durable over 2 months. In support of the importance of social experience and conversation for shaping children’s sociocognitive development (Carpendale and Lewis; 2004; Hughes, 2011), our results illustrate the feasibility and effectiveness of school-based interventions in which specific aspects of ToM—and their interpersonal and cognitive sequelae—can be fostered during the primary school years.

Acknowledgments We thank the children who took part in this study, their parents, and school staff members for their interest in and support of this project. We also thank Marta Nola for her support and contribution to the study.

Appendix A. Example item from the Strange Stories task (White et al., 2009) Jill wanted to buy a kitten, so she went to see Mrs. Smith, who had lots of kittens she didn’t want. Now Mrs. Smith loved the kittens, and she wouldn’t do anything to harm them, though she couldn’t keep them all herself. When Jill visited she wasn’t sure she wanted one of Mrs. Smith’s kittens since they were all males and she had wanted a female. But Mrs. Smith said, ‘‘If no one buys the kittens, I’ll just have to drown them!’’ QUESTION: Why did Mrs. Smith say that? 2 points—reference to persuasion, manipulating feelings, trying to induce guilt/pity; 1 point—reference to outcome (to sell them or get rid of them in a way that implies not drowning) or simple motivation (to make Jill sad); 0 points—reference to general knowledge without realization that the statement was not true (she’s a horrible woman)

Appendix B. Script of training procedures for each condition (1) ToM condition: Example of misunderstanding Story Question

—What was Robin intending —Right!/No, actually Robin to do? didn’t want to steal the cat. He wanted to do a good deed. He wanted to return it to the neighbor. —Why does the neighbor start to shout ‘‘Help me! Stop the thief!’’?

—What does Robin think about the neighbor’s behavior?

—Can Robin say or do something in order to stop his neighbor’s shouts? What? If he does this, why would the old woman stop shouting?

—Well done!/No, actually she shouts because she thinks he wants to steal her cat. She has misunderstood the situation. She has not understood Robin’s good intentions. —Right! / No, actually Robin imagines that the old woman didn’t understand what had happened because she forgot her glasses. —You’re right! / No, actually he could make his neighbor recognize him and explain that the cat was escaping and that he thought to run after it in order to return it. By doing this, the elderly woman would modify her point of view. She would realize her misunderstanding and she would understand what really happened.

Conversation onset provided by experimenter

Language exercise

Right! If Robin did what you said, the woman would stop shouting. She would recognize Robin and understand his good intentions about returning her cat. So she would change her point of view. Indeed, people’s beliefs can change, for example, when people understand that their ideas are wrong or that they have not gotten enough information in order to understand well. People can act or say things in order to change other people’s wrong beliefs. In this way, they can solve misunderstandings, just as Robin would do if he made his neighbor recognize him and he explained to her that he wanted to return the cat. Imagine a misunderstanding episode similar to Robin’s story. Describe it, explaining what you would do in that situation in order to solve it.

What is, in your opinion, the meaning of this sentence in the story? ‘‘he thinks to run after it’’: —he imagines to run after it —he decides to run after it⁄

—he believes to run after it

—he understands to run after it

S. Lecce et al. / Journal of Experimental Child Psychology 126 (2014) 52–67

It is evening time, and Robin is taking a bag of rubbish outside to put in the bin. Suddenly he sees his neighbor’s cat running away. He thinks to run after it in order to return it to his old neighbor. Luckily he manages to catch the cat. At that moment, his neighbor opens the door and glimpses her cat struggling in the arms of a boy. She has left her glasses in the dining room, so she can’t see well. She starts to shout, ‘‘Help me! Stop the cat thief!’’

Feedback example for each question

(continued on next page) 63

64

Appendix B (continued)

(1) ToM condition: Example of misunderstanding Story Question

(2) Control condition Story

Conversation onset provided by experimenter

Language exercise

Feedback example for each question

Conversation onset provided by experimenter

Language exercise

—What does Sarah do when —Okay / No, actually she writes Sometimes people are very busy and they can accidentally leave teachers give homework? the homework down in her an object somewhere. Later, diary. when they need that object, they can’t find it. So they can look for it starting in the last place in which they used it. —Are Sarah and her classmates always in the same place? If no, where do they go? (List the places in the right order.)

—Right answer! / No, actually they went to different places. They were in the English Have you ever lost an object? classroom first, then they went How did you go about finding to the art room, then to the gym, it? Tell your classmates about it. and finally to the computer lab.

—In which places did Sarah —Well done! / No, actually use the diary? Sarah used the diary in the English classroom and in the art room. —Why does the teacher look —You’re right! / The teacher for the diary in art room and doesn’t look for the diary in the English classroom because Sarah in the gym but not in the didn’t leave her diary there. She English classroom? used it in the next place she went, that is, the art room. She left her diary in the art room or in the gym.

Note. The researcher’s comments are given in italics. Asterisks indicate the right answer.

What is, in your opinion, the meaning of this sentence in the story? ‘‘teachers give homework’’: —teachers do homework —teachers set homework⁄ —teachers check homework —teachers have homework

S. Lecce et al. / Journal of Experimental Child Psychology 126 (2014) 52–67

Today is Friday. Teachers are giving some homework to their students for the weekend. Sarah always writes all her homework in her diary. During English class the teacher gives homework to the children, and Sarah writes it in her diary. Then students go to the art room. There the teacher says, ‘‘Write down that you must draw a picture of your family with crayons by next Monday.’’ After art class they go to the gym and then to computer lab. When the computer class teacher gives homework, Sarah can’t write it down because she can’t find her diary. So she tells the teacher. The teacher looks for the diary in the art room and in the gym, but not in the English classroom.

Question

Feedback example for each question

S. Lecce et al. / Journal of Experimental Child Psychology 126 (2014) 52–67

65

References Apperly, I. A. (2011). Mindreaders: The cognitive basis of ‘‘theory of mind’’. New York: Psychology Press. Appleton, M., & Reddy, V. (1996). Teaching three-year-olds to pass false belief tests: A conversational approach. Social Development, 5, 275–291. Banerjee, R., Watling, D., & Caputi, M. (2011). Peer relations and the understanding of faux pas: Longitudinal evidence for bidirectional associations. Child Development, 82, 1887–1905. Baron-Cohen, S., O’Riordan, M., Stone, V., Jones, R., & Plaisted, K. (1999). Recognition of faux pas by normally developing children and children with Asperger syndrome or high-functioning autism. Journal of Autism and Developmental Disorders, 29, 407–418. Blair, C., & Razza, R. P. (2007). Relating effortful control, executive function, and false belief understanding to emerging math and literacy ability in kindergarten. Child Development, 78, 647–663. Bull, R., & Scerif, G. (2001). Executive functioning as a predictor of children’s mathematics ability: Inhibition, switching, and working memory. Developmental Neuropsychology, 19, 273–293. Caputi, M., Lecce, S., Pagnin, A., & Banerjee, R. (2012). Longitudinal effects of theory of mind on later peer relations: The role of prosocial behavior. Developmental Psychology, 48, 257–270. Carlson, S. M., & Moses, L. J. (2001). Individual differences in inhibitory control and children’s theory of mind. Child Development, 72, 1032–1053. Carpendale, J. I. M., & Lewis, C. (2004). Constructing an understanding of mind: The development of children’s social understanding within social interaction. Behavioral and Brain Sciences, 27, 79–151. Cavallini, E., Lecce, S., Bottiroli, S., Palladino, P., & Pagnin, A. (2013). Beyond false belief: Theory of mind in young, young-old, and old-old adults. International Journal of Aging and Human Development, 76, 181–198. Clements, W., Rustin, C. L., & McCallum, S. (2000). Promoting the transition from implicit to explicit understanding: A training study of false belief. Developmental Science, 3, 81–92. Cornoldi, C., & Colpo, G. (1998). Prove di lettura MT per la scuola elementare 2 [MT Tests of Reading for Primary School 2]. Firenze, Italy: Organizzazioni Speciali. Cutting, A. L., & Dunn, J. (1999). Theory of mind, emotion understanding, language, and family background: Individual differences and interrelations. Child Development, 70, 853–865. Cutting, A. L., & Dunn, J. (2006). Conversations with siblings and with friends: Links between relationship quality and social understanding. British Journal of Developmental Psychology, 24, 73–87. Davidson, M., Amso, D., Anderson, L. C., & Diamond, A. (2006). Development of cognitive control and executive function from 4 to 13 years: Evidence from manipulations of memory, inhibition, and task switching. Neuropsychologia, 44, 2037–2078. de Rosnay, M., & Hughes, C. (2006). Conversation and theory of mind: Do children talk their way to socio-cognitive understanding? British Journal of Developmental Psychology, 24, 7–37. Devine, R. T., & Hughes, C. (2013). Silent films and strange stories: Theory of mind, gender, and social experiences in middle childhood. Child Development, 84, 989–1003. Dunn, J. (1996). Children’s relationships: Bridging the divide between cognitive and social development. Journal of Child Psychology and Psychiatry and Allied Disciplines, 37, 507–518. Dunn, J., & Brophy, M. (2005). Communication, relationships, and individual differences in children’s understanding of mind. In J. W. Astington & J. A. Baird (Eds.), Why language matters for theory of mind (pp. 50–69). Oxford, UK: Oxford University Press. Dunn, J., Brown, J., & Beardsall, L. (1991). Family talk about feeling states and children’s later understanding of others’ emotions. Developmental Psychology, 27, 448–455. Dyer, J. R., Shatz, M., & Wellman, H. M. (2000). Young children’s storybooks as a source of mental state information. Cognitive Development, 15, 17–37. Ensor, R., Devine, R. T., Marks, A., & Hughes, C. (2014). Mothers’ cognitive references to two-year-olds predict theory of mind at ages 6 and 10. Child Development, 85, 1222–1235. Ensor, R., & Hughes, C. (2008). Content or connectedness? Mother–child talk and early social understanding. Child Development, 79, 201–216. Espy, K. A., McDiarmid, M. M., Cwik, M. F., Stalets, M. M., Hamby, A., & Senn, T. E. (2004). The contribution of executive functions to emergent mathematic skills in preschool children. Developmental Neuropsychology, 26, 465–486. Filippova, E., & Astington, J. W. (2008). Further development in social reasoning revealed in discourse irony understanding. Child Development, 79, 126–138. Flavell, J. H. (1993). The development of children’s understanding of false belief and the appearance–reality distinction. International Journal of Psychology, 28, 595–604. Flay, B. R., Biglan, A., Boruch, R. F., Castro, F., Gottfredson, D., Kellam, S., et al (2005). Standards of evidence: Criteria for efficacy, effectiveness, and dissemination. Prevention Science, 6, 151–175. Gopnik, A., & Astington, J. W. (1988). Children’s understanding of representational change and its relation to the understanding of false belief and the appearance–reality distinction. Child Development, 59, 26–37. Grazzani, I., & Ornaghi, V. (2012). How do use and comprehension of mental-state language relate to theory of mind in middle childhood? Cognitive Development, 27, 99–111. Hale, C. M., & Tager-Flusberg, H. (2003). The influence of language on theory of mind: A training study. Developmental Science, 6, 346–359. Happé, F. G. E. (1994). An advanced test of theory of mind: Understanding of story characters’ thoughts and feelings by able autistic, mentally handicapped, and normal children and adults. Journal of Autism and Developmental Disorders, 24, 129–154. Harris, P. (1999). Acquiring the art of conversation: Children’s developing conception of their conversation partner. In M. Bennett (Ed.), Developmental psychology: Achievements and prospects (pp. 89–105). London: Psychology Press. Hughes, C. (1998a). Executive function in preschoolers: Links with theory of mind and verbal ability. British Journal of Developmental Psychology, 16, 233–253. Hughes, C. (1998b). Finding your marbles: Does preschoolers’ strategic behavior predict later understanding of mind? Developmental Psychology, 34, 1326–1339.

66

S. Lecce et al. / Journal of Experimental Child Psychology 126 (2014) 52–67

Hughes, C. (2011). Social understanding and social lives: From toddlerhood through to the transition to school. London: Psychology Press. Hughes, C., & Cutting, A. (1999). Nature, nurture, and individual differences in early understanding of mind. Psychological Science, 10, 429–432. Hughes, C., Devine, R., Ensor, R., Koyasu, M., Misokawa, A., & Lecce, S. (2014). Lost in translation? Comparing British, Japanese, and Italian children’s theory of mind performance. Child Development Research, 2014, 1–10. Hughes, C., Jaffee, S. R., Happé, F., Taylor, A., Caspi, A., & Moffitt, T. E. (2005). Origins of individual differences in theory of mind: From nature to nurture? Child Development, 76, 356–370. Hughes, C., Marks, A., Ensor, R., & Lecce, S. (2010). A longitudinal study of conflict and inner state talk in children’s conversations with mothers and younger siblings. Social Development, 19, 822–837. Huizinga, M., Dolan, C. V., & van der Molen, M. W. (2006). Age-related change in executive function: Developmental trends and a latent variable analysis. Neuropsychologia, 44, 2017–2036. Kloo, D., & Perner, J. (2008). Training theory of mind and executive control: A tool for improving school achievement? Mind, Brain, and Education, 2, 122–127. Lecce, S., Caputi, M., & Hughes, C. (2011). Does sensitivity to criticism mediate the relationship between theory of mind and academic achievement? Journal of Experimental Child Psychology, 110, 313–331. Lecce, S., Caputi, M., & Pagnin, A. (2014). Long-term effect of theory of mind on academic achievement: The role of sensitivity to criticism. European Journal of Developmental Psychology, 11, 305–318. Lecce, S., & Hughes, C. (2010). ‘‘The Italian job’’? Comparing theory of mind performance in British and Italian children. British Journal of Developmental Psychology, 28, 747–766. Lecce, S., & Hughes, C. (2014). Reading minds and stories: Theory of mind predicts reading in middle childhood, in preparation. Lecce, S., Zocchi, S., Pagnin, A., Palladino, P., & Taumoepeau, M. (2010). Reading minds: The relation between children’s mental state knowledge and their metaknowledge about reading. Child Development, 81, 1876–1893. Lockl, K., & Schneider, W. (2007). Knowledge about the mind: Links between theory of mind and later metamemory. Child Development, 78, 148–167. Lohmann, H., & Tomasello, M. (2003). The role of language in the development of false belief understanding: A training study. Child Development, 74, 1130–1144. Meins, E., Fernyhough, C., Johnson, F., & Lidstone, J. (2006). Mind-mindedness in children: Individual differences in internalstate talk in middle childhood. British Journal of Developmental Psychology, 24, 181–196. Melot, A. N., & Angeard, N. (2003). Theory of mind: Is training contagious? Developmental Science, 6, 178–184. Meristo, M., Falkman, K. W., Hjelmquist, E., Tedoldi, M., Surian, L., & Siegal, M. (2007). Language access and theory of mind reasoning: Evidence from deaf children in bilingual and oralist environments. Developmental Psychology, 43, 1156–1169. Miller, S. A. (2009). Children’s understanding of second-order mental states. Psychological Bulletin, 135, 749–773. Miller, S. A. (2012). Theory of mind beyond the preschool years. New York: Psychology Press. Milligan, K., Astington, J., & Dack, L. A. (2007). Language and theory of mind: Meta-analysis of the relation between language ability and false-belief understanding. Child Development, 78, 622–646. Nelson, K. (2005). Language pathways into the community of minds. In J. W. Astington & J. A. Baird (Eds.), Why language matters for theory of mind (pp. 26–49). Oxford, UK: Oxford University Press. Nelson, K. (2007). Young minds in social worlds: Experience, meaning, and memory. Cambridge, MA: Harvard University Press. Olson, D. R., Antonietti, A., Liverta Sempio, O., & Marchetti, A. (2006). The mental verbs in different conceptual domains and in different cultures. In A. Antonietti, O. Liverta Sempio, & A. Marchetti (Eds.). Theory of mind and language in different developmental contexts (pp. 31–64). New York: Springer. Ornaghi, V., Brockmeier, J., & Grazzani, I. (2011). The role of language games in children’s understanding of mental states: A training study. Journal of Cognition and Development, 12, 239–259. Ornaghi, V., Brockmeier, J., & Grazzani, I. (2014). Enhancing social cognition by training children in emotion understanding: A primary school study. Journal of Experimental Child Psychology, 119, 26–39. Orsini, A. (1997). WISC-R: Contributo alla taratura italiana [WISC-R: Contribution to Italian norms]. Firenze, Italy: Organizzazioni Speciali. Peterson, C. C., & Siegal, M. (1995). Deafness, conversation, and theory of mind. Journal of Child Psychology and Psychiatry and Allied Disciplines, 36, 459–474. Peterson, C. C., & Slaughter, V. (2003). Opening windows into the mind: Mothers’ preferences for mental state explanations and children’s theory of mind. Cognitive Development, 18, 399–429. Peterson, C. C., & Slaughter, V. (2006). Telling the story of theory of mind: Deaf and hearing children’s narratives and mental state understanding. British Journal of Developmental Psychology, 24, 151–179. Pillow, B. H. (1991). Children’s understanding of biased social cognition. Developmental Psychology, 27, 539–551. Ronald, A., Viding, E., Happé, F., & Plomin, R. (2006). Individual differences in theory of mind ability in middle childhood and links with verbal ability and autistic traits: A twin study. Social Neuroscience, 1, 412–425. Ruffman, T., Slade, L., & Crowe, E. (2002). The relation between children’s and mother’s mental state language and theory-ofmind understanding. Child Development, 73, 734–751. Shallice, T. (1982). Specific impairments of planning. Philosophical Transactions of the Royal Society of London B, 298, 199–209. Slaughter, V., & Gopnik, A. (1996). Conceptual coherence in the child’s theory of mind: Training children to understand belief. Child Development, 67, 2967–2988. Tomasuolo, E., Valeri, G., Di Renzo, A., Pasqualetti, P., & Volterra, V. (2013). Deaf children attending different school environments: Sign language abilities and theory of mind. Journal of Deaf Studies and Deaf Education, 18, 12–29. Torsheim, T., Boyce, W., Currie, C., & Zambon, A. (2006). The family affluence scale as a measure of national wealth: Validation of an adolescent self-report measure. Social Indicators Research, 78, 473–487. Turnbull, W., & Carpendale, J. M. (1999). A social pragmatic model of talk: Implications for research on the development of children’s social understanding. Human Development, 42, 328–355. Wellman, H. M., Cross, D., & Watson, J. (2001). Meta-analysis of theory-of-mind development: The truth about false belief. Child Development, 72, 655–684.

S. Lecce et al. / Journal of Experimental Child Psychology 126 (2014) 52–67

67

White, S., Hill, E., Happé, F. G. E., & Frith, U. (2009). Revisiting the strange stories: Revealing mentalizing impairments in autism. Child Development, 80, 1097–1117. Ziv, M., Smadja, M. L., & Aram, D. (2013). Mothers’ mental-state discourse with preschoolers during storybook reading and wordless storybook telling. Early Childhood Research Quarterly, 28, 177–186.