copy the linklink copied!Chapter 5. Children’s social-emotional skills in Estonia
This chapter presents findings on the social-emotional skills of five-year-olds in Estonia. It shows the differences in social-emotional scores across multiple subgroups of children, considering their individual and family characteristics, as well as their home learning environments. This is based on a direct assessment of children’s skills and reports from the children’s parents and educators.
copy the linklink copied!The importance of social-emotional development
Children develop their capacity to experience and express emotions starting in early infancy, at the same time as they grow physically and cognitively in developing their language and problem-solving skills (Thompson, 2001[1]). Recent developments in neuroscience have shown that the same neural circuits involved in the regulation of emotions overlap with those associated with cognitive processing (Bush, Luu and Posner, 2000[2]; Davidson et al., 2002[3]; Posner and Rothbart, 2000[4]).
Emotions can support cognitive development when they are well-regulated, but that they interfere when they are not. For instance, children who do not feel in control of their emotions are more prone to outbursts, inattention and rapid retreats from stressful situations (Garber and Dodge, 1991[5]). Children’s beliefs and their neural mechanisms of attention are interrelated components during childhood development (Schroder et al., 2017[6]).
Early social-emotional skills are strong predictors of later health, educational, social and labour-market outcomes
Understanding emotions is a unique, concurrent predictor of academic competence (Leerkes et al., 2008[7]). Early prosocial behaviour at age eight is shown to be as important as early cognitive ability in predicting educational attainment at age 30 (Schoon et al., 2015[8]), as well as in shaping attainment in adolescence and adulthood (Caprara et al., 2000[9]). Social-emotional skills developed during childhood are linked to educational achievement, even after controlling for early literacy and numeracy skills (Duncan et al., 2007[10]). For example, children’s early skills in identifying and responding empathetically to others’ emotions have been found to predict concept knowledge and language competence, even after controlling for age, gender and parental income level (Rhoades et al., 2011[11]; Garner and Waajid, 2008[12]).
Underdeveloped skills in identifying others’ emotions in early adolescence predict increases in fear, decreases in positive emotions and decreases in the quality and quantity of social support. Amongst boys, low emotion identification skills also predict increases in sadness (Ciarrochi, Heaven and Supavadeeprasit, 2008[13]).
Early empathy, trust and prosocial behaviours are associated with social justice beliefs and a lower likelihood of involvement in crime and delinquency in adulthood (Schoon et al., 2015[8]). Low empathy is associated with antisocial and delinquent behaviours, and increased risk of psychopathology as adults (Fontaine et al., 2011[14]). Sympathy and moral reasoning among six- to nine-year-olds are associated with social justice values at age 12 (Daniel et al., 2014[15]).
Children’s emotional health is the strongest predictor of adult life satisfaction at all ages, even more than family economic resources, family psychosocial resources and children’s cognitive ability (Flèche, Lekfuangfu and Clark, 2019[16]). Early emotional well-being is linked with mental health in later life, and emotional difficulties at age five are predictors of midlife psychological disorders, such as anxiety and depression (Buchanan, Flouri and Brinke, 2002[17]; Rutter, Kim-Cohen and Maughan, 2006[18]).
IELS included a direct measure of children’s emotion identification and attribution, and indirect measures of children’s prosocial behaviour, trust in familiar people and non-disruptive behaviour
The International Early Learning and Child Well-being Study (IELS) provides a direct and indirect assessment of social-emotional skills (Box 5.1). Parents and educators responded to survey questions about the child’s prosocial behaviour, trust and disruptive behaviours. Children in the study participated in an interactive tablet-based assessment of their empathy skills in a one-on-one setting with a trained study administrator. Reports from educators and parents helped to create a fuller picture of children’s early social-emotional skills in both home and early childhood education and care (ECEC) environments than could be ascertained from the direct assessment alone.
Social-emotional learning is the process through which children and adults acquire and effectively apply the knowledge, attitudes, and skills necessary to understand and manage emotions; set and achieve positive goals; feel and show empathy for and towards others; establish and maintain positive relationships; and make responsible decisions (CASEL, 2015[22]; Wessberg et al., 2015[23]).
Social-emotional development is the continuous process of learning social-emotional skills. Similar to literacy and numeracy, developing these skills early on and continuing throughout adulthood is important for their effect on personal, academic and life outcomes over time.
Social-emotional skills are individual characteristics that 1) link biological predispositions and environmental factors; 2) are expressed through consistent patterns of thoughts, feelings and behaviours; 3) develop through formal and informal learning experiences; and 4) influence important socio-economic outcomes throughout life (De Fruyt and Wille, 2015[24]). The term is increasingly prevalent in policy discussions that emphasise improving these skills through learning. Other terms such as “21st century skills”, “non-cognitive skills”, “employability skills” and “personality characteristics” often refer to the same concept. For further discussion about their overlaps and differences, see Abrahams et al. (2019[25]) and Kankaraš and Suarez-Alvarez (2019[26]).
Measuring empathy entails the assessment of two skills: emotion identification and emotion attribution in response to a story about a set of characters. Children who participated in the IELS direct assessment responded to hypothetical (story) scenarios designed to measure their empathy skills. Narrated stimulus stories presented cartoon-like children in brief vignettes presented on electronic tablets. The empathy measure required the child to identify an emotion using emoticons representing happy, sad, afraid, angry and surprised. The emotion identification scores reflected children’s ability to recognise the emotions of others (i.e. how did the story character feel?). The emotion attribution scores reflected the interaction of concordant emotional response (i.e. when child’s responses matched the emotion of the story character) and his or her own emotion attribution (i.e. how the child felt and why s/he felt that way in response to the story).
IELS also measured prosocial, trust and non-disruptive behaviours indirectly through reports from parents and educators, with parents and educators rating the same children on the same set of behaviours. The items for assessing prosocial behaviour and non-disruptive behaviour were based on the Adaptive Social Behaviour Inventory (Hogan, Scott and Bauer, 1992[19]), while those for trust were developed based on previous research (Baumrind, 1968[20]; Roberts, Strayer and Denham, 2014[21]). The prosocial behaviour measure is composed of items such as the child “understands others’ feelings, such as when they are happy, sad or angry”. The non-disruptive behaviour measure was composed of items such as the child “fights with other children”, which was positively inverted for easier interpretation (i.e. the higher the scores the less disruptive). Lastly, the trust measure is composed of items such as the child “approaches familiar adults for comfort when upset”.
This chapter compares educators and parents’ ratings of children’s behaviours related to their social-emotional skills. Parents undoubtedly have a better knowledge of their child in a wider set of situations, while educators have a larger reference group for comparison, but children may also behave differently in different environments.
Educators’ ratings on children’s behaviours were more closely related to the direct assessment of social-emotional skills and their scores were aggregated into a single score for prosocial behaviour, trust and non-disruptive behaviour, and scaled together with the rest of the study’s outcomes. Educators’ indirect assessments are, therefore, internationally standardised with a mean of 500 and a standard deviation of 100, and comparable with the scores from other subdomains of the children’s direct assessment.
IELS measures of social-emotional skills are interrelated
An important component of prosocial behaviour and getting along with others is being able to recognise and understand the emotions of others (Strayer, 1987[27]; Strayer, 1993[28]). Both emotion identification and emotion attribution act, therefore, as precursors to engaging in prosocial behaviour in response to another person’s emotional state (Hinnant and O’Brien, 2007[29]). At the same time, it is important to note that prosocial behaviour goes one-step further as it also includes the expression of positive social behaviours, such as: the child “tries to comfort others when they are upset”.
The central aspect of trust in IELS is the child’s expectations that others will be supportive, responsive and kind (Bowlby, 1983[30]). Children develop their first relationships with adults, peers and friends in early childhood. When these first relationships are consistent, predictable and responsive to their needs, children are more likely to develop secure attachments that help them to acquire and reinforce their trust in known people and themselves (Bowlby, 1983[30]). It is important to clarify that trust does not mean that children are indiscriminately developing secure attachments with anybody without judgement, but that they develop trust because of frequent and repetitive patterns with close adults. Reassuring expressions from caregivers (which nurture a child’s secure attachment) can support children to continue to play comfortably, while anxious expressions (which nurture a child’s insecure attachments) might interfere in children’s trust and playful interactions and, ultimately, hamper their development (Baldwin and Moses, 1996[31]). Mistrustful children might be overly wary or fearful of peers or adults; a child might be reluctant to engage with others, or be needy and dependent since s/he does not trust others to be responsive and supportive. As shown in this chapter, children’s trust is associated with adaptive social behaviour, such as the expression of prosocial and non-disruptive behaviour.
copy the linklink copied!Social-emotional skills of five-year-olds in Estonia
The average five-year-old child in Estonia exhibit social-emotional skills at the same or higher levels than their counterparts in England and the United States, although educators in Estonia rate five-year-olds as more disruptive than educators in England or the United States rated their respective children.
When presented with a range of stories and situations, children in Estonia were better able to identify the feelings of the characters than children in the United States and England. The mean for five-year-olds in Estonia on emotion identification was 511 points, which is significantly higher than England (497) and the United States (493). Children’s ability to recognise emotions is a precursor of their ability to feel empathy for others. In emotion attribution, where the score reflects children’s own emotions, children in Estonia scored similar to those in the United States and England.
According to educators, children in Estonia had significantly higher prosocial behaviour (511) than children in the United States (494) and England (495). However, educators in Estonia (470) rated children significantly more disruptive than children in the United States (515) and England (515). Educators in the three countries participating in the study rated children’s average levels of trust similarly.
Scores were scaled so that the overall mean in each domain (emotion identification, emotion attribution, prosocial behaviour, trust and non-disruptive) was 500, and the standard deviation 100. Each country contributed equally to the computation of that mean. The distributions of social-emotional scores in Estonia are shown in Figure 5.1.
Note: Graphs produced using the first plausible values.
Social-emotional learning scores are interrelated in both the direct and indirect assessments
Table 5.1 shows the correlation coefficients between the social-emotional skills measured as part of IELS for Estonia. For the direct assessment, the scores for emotion identification and emotion attribution were strongly correlated (r=0.54). For the indirect assessments (both educators and parents), the association between prosocial behaviour and non-disruptive behaviour was strong. The strongest association was between ratings of prosocial behaviour and trust. The association between trust and non-disruptive behaviour in Estonia was weaker than in the United States and England. As expected, these results were similar to the overall correlations across participating countries in IELS.
The association between the direct assessment of children and educators’ indirect assessment is moderately strong. The direct assessment provides children’s scores on emotion identification and emotion attribution, while the indirect assessment provides educator views on children’s prosocial behaviour, trust and non-disruptive behaviour. Examples of prosocial behaviour include the child “understands others’ feelings” and “tries to comfort others when they are upset”. While the first statement is closely associated with the tasks in the direct assessment, the second statement includes a positive behaviour. Examples of trust include the child “approaches familiar adults for comfort when upset” and for disruptive behaviours: the child “fights with other children”. Although these behaviours still relate to the tasks presented in the direct assessment, they are slightly more distal behaviours from emotion identification and emotion attribution than prosocial behaviour.
On the other hand, the association between educators’ and parents’ indirect assessments is moderate while the association between parents’ ratings and the direct assessment of children’s social and emotional is smaller. As previously mentioned, these domains are conceptually overlapping, but not exactly the same.
Parents give more positive ratings of their children’s empathy skills than educators but rate children’s emotional control as less developed
In addition to the direct assessment of emotion identification and emotion attribution, parents and educators rated children’s development in empathy (e.g. the child is considerate, helpful, caring) and emotional control (e.g. the child controls emotions, waits patiently for something he or she wants). Parents were more likely to rate children’s empathy skills as more developed than educators (Figure 5.2). However, parents rated children’s emotional control as less developed than educators. Parents in England and the United States also rated children’s empathy skills as more developed than educators. However, educators in England and the United States rated children’s emotional control similarly to parents.
Note: The figure is comparing the same children rated by their educators and parents.
copy the linklink copied!Individual characteristics and social-emotional skills
Girls typically have better social-emotional outcomes than boys
Figure 5.3 shows that, on average, girls had better social-emotional scores than boys with respect to emotion identification, emotion attribution, prosocial behaviour, trust and non-disruptive behaviour. Educators reported higher gender differences in prosocial behaviour than were found in the direct assessment. The differences in scores between boys and girls were statistically significant for the direct and indirect assessments. The gender gap in Estonia is significantly larger than in the United States in prosocial behaviour and trust. However, the gender gap in Estonia is not significantly larger than in the United States in emotion identification, emotion attribution, and non-disruptive behaviour or in any of the five social-emotional domains in comparison to England.
Note: The gender differences in mean scores are statistically significant.
Both parents and educators reported girls as having more developed empathy and emotional control than boys (Figure 5.4). This difference also existed in England and the United States. Parents were also more likely than educators to rate children’s empathy skills as better developed regardless of their gender.
Note: The figure is comparing the same children rated by their educators and parents.
The gender gap in social-emotional skills is larger among Russian-speaking children
Figure 5.5 shows the differences between girls and boys by language (i.e. Estonian-speaking children and Russian-speaking children), with girls having, on average, better social-emotional outcomes than boys across both subgroups. However, Russian-speaking children had the highest gender difference in social-emotional scores. In prosocial behaviour, where the highest gap was found, Russian-speaking girls scored 66 points higher than Russian-speaking boys, compared to Estonian-speaking girls who scored 41 points higher than Estonian-speaking boys.
Note: The differences shown in this figure are all statistically significant.
At the same time, Russian-speaking girls scored on average 18 points more than Estonian-speaking girls across all social-emotional skills, , while this difference was 9 points for boys. The following section will provide further analyses of the differences in findings between Estonian and Russian-speaking children.
Children’s social-emotional skills scores increase slightly with age
In Estonia, the average difference between the oldest and youngest children was 108 points for emotion identification and 71 points for emotion attribution. This means, for every additional month in age, children’s emotion identification scores increased by 7 points on average and their emotion attribution scores by 4 points. On average, the oldest children in Estonia (6 years 0 months) scored approximately 80 points more than the youngest children (5 years 0 months) in the direct assessment domains. This means that a score-point difference of 20 points in this report roughly equates to a quarter year in the development of an average 5-years old child in these domains, assuming his/her development is approximately linear.
Figure 5.6 shows the social-emotional scores by children’s age in months at the time of assessment. The data indicate a small but significant positive correlation between children’s age and their scores on the direct assessment of their social-emotional learning. In Estonia, the correlation was 0.21 for emotion identification and 0.12 for emotion attribution. Differences by age were smaller in educator indirect assessment ratings: the correlation was significant for prosocial behaviour and not statistically significant for trust and non-disruptive behaviour. The data show similar correlations between age and social-emotional learning outcomes for boys and girls.
Social, emotional or behavioural difficulties are more strongly associated with lower social-emotional skills, especially more disruptive behaviour, than low birth weight or learning difficulties
Parents in IELS provided information on whether their child had a low birth weight (under 2.5 kg) or premature birth, learning difficulties, or social, emotional or behavioural difficulties. The child’s ECEC centre also provided information on whether they were classified as having special educational needs (SEN).
The cause of low birth weight is often premature birth (i.e. born before 37 weeks gestation). Parents of 8% of 5-year-old children in Estonia reported that their children had low birth weight or premature birth, slightly lower to the share in England (11%) and the United States (10%). Children with learning difficulties (e.g. speech or language delay, intellectual disability) in Estonia represented 10% of children, which was 3 percentage points lower than in the United States, and similar to England. Children with emotional difficulties (e.g. social, emotional or behavioural difficulties) represented 10% of children in Estonia, which was approximately 2 percentage points higher than in England and lower than in the United States. Children with SEN (e.g. cognitive, behavioural or emotional disability) represented 5% of children in Estonia, which was less 2 percentage points lower than in England (the United States had no available data).
In Estonia, boys were more likely than girls to be identified by their parents as having learning difficulties (13% for boys and 7% for girls); having social, emotional or behavioural difficulties (14% for boys and 7% for girls); and having SEN (6% of boys and 3% of girls). However, the data showed no significant gender differences in social-emotional learning scores between children with and without these difficulties, after accounting for socio-economic status (the low number of children with SEN do not allow a robust analysis by gender).
Overall, 17% of children in Estonia had experienced one of these three challenges, as reported by parents, 4% had experienced two and fewer than 1% had experienced all three. In other words, 78% of parents reported that their children had experienced none of these challenges, which is similar to the percentage in England and the United States.
Children who had experienced learning difficulties had lower emotion identification and emotion attribution skills, and were rated by their educators as having lower prosocial behaviour and trust, than children without learning difficulties, after controlling for socio-economic status. Children with social, emotional or behavioural difficulties had lower emotion identification, prosocial behaviour and trust, as well as more disruptive behaviour, after controlling for socio-economic status.
When all of these challenges were analysed together, social, emotional or behavioural difficulties were more highly associated with poor social-emotional outcomes than low birth weight, or learning difficulties – except in the case of emotion identification and trust, where learning difficulties were also highly associated with poorer outcomes. Disruptive behaviour was particularly associated with social, emotional or behavioural difficulties, as might be expected. These associations were significant after controlling for socio-economic status.
When analysed alone, children with a low birth weight or premature birth had lower emotion identification skills than children born weighing over 2.5 kg. However, these differences disappeared when controlling for children’s learning and social, emotional or behavioural difficulties.
Note: Darker coloured markers indicate that the difference is statistically significant.
copy the linklink copied!Home and family characteristics and early social-emotional skills
Children from advantaged socio-economic backgrounds in Estonia have higher social-emotional scores, but the differences were narrower than in the other two countries participating in the study
Figure 5.8 shows the difference in social-emotional learning outcomes between children from the top and bottom quartile of the national socio-economic status (SES) index1. IELS defines children from an advantaged socio-economic background as those located in the top quartile of socio-economic status. Children from a disadvantaged socio-economic background are defined as those located in the bottom quartile. The results show that children from advantaged socio-economic backgrounds had higher outcomes than children from disadvantaged socio-economic backgrounds in both direct and indirect assessments.
Note: The differences in the mean scores of children in the bottom and top quartiles of socio-economic status are statistically significant.
Nevertheless, the strength of the relationship varied depending on the social-emotional outcome. The direct assessment showed that socio-economic status had a significant relationship with emotion identification, but not on emotion attribution. According to educators, children from an advantaged socio-economic background had higher prosocial behaviour and trust, but similar disruptive behaviour as children from other socio-economic backgrounds.
In Estonia, socio-economic status was significantly positively correlated with emotion identification (r = 0.17), prosocial behaviour (r = 0.16), and trust (r = 0.14), and not significantly associated with emotion attribution and disruptive behaviour. In the United States, socio-economic status was significantly positively correlated with emotion identification (r = 0.18), emotion attribution (r = 0.08), prosocial behaviour (r = 0.15), trust (r = 0.12), and not significantly associated with disruptive behaviour. In England, socio-economic status was significantly positively correlated with emotion identification (r = 0.15), emotion attribution (r = 0.17), prosocial behaviour (r = 0.17), trust (r = 0.11), and non-disruptive behaviour (r = 0.10).
Some family characteristics are not associated with the social-emotional skills of children.
In Estonia, around 6% of children lived in homes where at least one parent mostly spoke a language other than the assessment language, which is lower than in England (16%) and the United States (20%). In Estonia, the two languages of assessment were Estonian and Russian. The results from both direct and indirect assessments showed no significant differences in the results of children who spoke a language at home other than Estonian or Russian and the results of other children, either before or after controlling for socio-economic status.
In Estonia, 12% of children lived in single-parent households, which was lower than in England (15%) and the United States (15%). The direct and indirect assessments showed no statistically significant differences between one- and two-parent households in terms of children’s social-emotional outcomes, either before or after controlling for socio-economic status. Parents’ age was also not significantly associated with social-emotional skills after accounting for socio-economic background.
IELS defines an immigrant background as having both parents – or the sole parent if a single parent – born in another country or economy than where the study took place. In Estonia, less than 2% of children had an immigrant background, which is substantially lower than in the United States (18%) and England (18%). The low number of children with an immigrant background in Estonia was not sufficient to provide robust comparative results.
Russian-speaking children have better social-emotional outcomes than Estonian-speaking children
In Estonia, Russian-speaking children represented 21% of the children. Figure 5.9 shows the social-emotional scores of Estonian-speaking children and Russian-speaking children, before and after controlling for socio-economic status. Russian-speaking children had significantly higher outcomes than Estonian-speaking children in emotion identification, emotion attribution, prosocial behaviour and trust. According to educators, however, Russian-speaking children were as disruptive as Estonian-speaking children.
Note: Darker coloured markers indicate that the difference is statistically significant. Russian-speaking children refer to children who took the assessment in Russian.
As previously noted, Russian-speaking children showed more prominent gender differences than Estonian-speaking children. However, Russian-speaking girls did not significantly contribute more than boys to the overall differences, after accounting for socio-economic status and the overall gender differences on outcomes.
Children whose mothers had completed tertiary education have better social-emotional outcomes than those whose mothers had not
In Estonia, 53% of mothers with five-year-olds had completed tertiary education (i.e. bachelor’s degree or master’s degree, professional degree or doctorate), which was higher than in the other two countries participating in the study (40% in England and 39% in the United States). Figure 5.10 shows the social-emotional scores of children whose mothers had completed tertiary education and those whose mothers had not, with results showing that children whose mothers had completed tertiary education had higher emotion identification and higher prosocial behaviour and trust, after accounting for household income. Maternal education is also positively associated with social-emotional scores after accounting household income in England and the United States.
Note: All differences are statistically significant.
Being an only-child is associated with higher disruptive behaviour and lower emotion identification
In Estonia, approximately 20% of children had no siblings, 50% had one, 21% had two, 6% had three, 1% had four and around 1% had five or more. Across participating countries, most children had one sibling. On average, children in Estonia had a lower number of siblings than in the other two countries: around 8% of children had three or more siblings compared to 12% in England and 20% in the United States.
The average 5-years old child with no siblings in Estonia had higher disruptive behaviour scores than children with one or more siblings and lower mean emotion identification scores than children with one sibling (Figure 5.11). These differences remained significant after controlling for socio-economic background. Children without siblings were also reported by educators in England as having higher disruptive behaviour than children with one or more siblings. However, this result was not found in the United States after controlling for socio-economic status.
copy the linklink copied!Home learning environment and early social-emotional skills
The number of books children have access to at home is positively related to their social-emotional skills
In Estonia, 13% of children lived in households with 10 children’s books or fewer, 23% in homes with 11-25 books, 32% in homes with 26-50 books, 21% in homes with 51-100 books and 10% in homes with more than 100 children’s books. Children in Estonia had, on average, a lower number of children’s books than children in the United States and England (the most frequent category in both the United States and England was more than 100). Children in homes with more books had, on average, higher social-emotional outcomes in the direct assessment of emotion identification and emotion attribution. This association remained significant after controlling for socio-economic status (Figure 5.12).
The indirect assessment yielded similar findings. Educators reported significantly higher prosocial behaviour and less disruptive behaviour in children from homes with more than 100 children’s books, in comparison with households that had between 26 and 50 books, after accounting for socio-economic status. The positive association of having a higher number of books at home did not significantly differ by gender, after accounting for socio-economic status.
Role-playing and having back-and-forth conversations with parents is associated with social-emotional skills, but doing it every day is not always better.
In Estonia, 6% of children never role-played with their parents (defined as imaginative or pretend play such as playing the role of a chef or shopkeeper), 28% did so less than once a week, 37% did so one or two days a week, 21% did so between three and four days a week, and 9% did so between five and seven days a week. The percentage of children in Estonia who role-played with their parents between three and seven days a week (30%) was lower than in the other two countries participating in the study (around 60% in the United States and 50% in England).
Children who role-played with their parents once or twice a week were better able to recognise emotions and were rated by their educators as having more trust than children who did so between five and seven days a week. (Figure 5.13). However, children who did these activities between five and seven days a week were less disruptive. These results remained significant after controlling for socio-economic status. The results did not significantly differ by gender, after accounting for socio-economic status. While these results were not significant for emergent literacy or numeracy, there was a positive association between children who role-played one or two days a week and scores in working memory compared to children who never role-played, after accounting for socio-economic status.
Role-playing with parents also showed differences in children’s social-emotional scores in the other two countries, but the frequency was differently associated with the outcomes. In England, children who did role-play with their parents one or two days a week had significantly higher emotion attribution than children who did not. In the United States, children who did role-play with their parents between 5 and 7 days a week had significantly higher mean emotion identification, emotion attribution, and prosocial behaviour scores than children who did it less than once a week.
In Estonia, fewer than 1% of parents reported that they never had back-and-forth conversations with their children about how they feel, 5% did so less than once a week, 17% one or two days a week, 29% three or four days a week, and 48% between five and seven days a week. These percentages were similar to England and United States. Children who had back-and-forth conversations with their parents between five and seven days a week were less disruptive, according to their educators, than those who did so between three and four days a week, after accounting for socio-economic status. Children in England and the United States who regularly have back-and-forth conversations about how they feel also had higher mean social-emotional scores.
Children whose parents read to them regularly have higher mean emotion identification, prosocial behaviour and trust scores.
In Estonia, 17% of parents read from a book to their child less than once a week, 24% did so one or two days a week, 22% three or four days a week, and 38% between five and seven days a week. The percentage of parents who read from a book to their child between five and seven days a week was lower than in the United States (43%) and England (59%). Children whose parents read to them between five and seven days a week had significantly higher mean trust scores than those whose parents read to them only one or two days a week. At the same time, children whose parents read to them less than one day a week had significantly lower mean emotion identification and prosocial behaviour scores than those whose parents read to them 3 or 4 days a week or between 5 to 7 days a week respectively, after controlling for socio-economic status (Figure 5.14). This association did not significantly differ by gender, after accounting for socio-economic status.
Children who regularly attended activities outside of the home are more empathetic and have stronger prosocial behaviours and trust than those who attended once a week or less
In Estonia, 22% of parents never took their child to special activities outside of the home, 14% did so less than once a week, 40% one or two days a week, 20% three or four days a week, and 5% between five and seven days a week. Examples of special activities include sports clubs, dance, swimming lessons or language lessons. The percentage of parents who took their child to activities outside of the home between one and two days a week was 36% in the United States and 47% in England. Children who attended special activities outside of the home between 5 and 7 days a week had higher mean empathy scores than those who did so one or two days a week, and were also rated as having stronger prosocial behaviour and trust by their educators than those who went one or two days a week or less, after accounting for socio-economic status (Figure 5.15). The positive relationship between going to special activities outside of the home one or two days a week and social-emotional scores did not significantly differ by gender.
According to educators, children whose parents were very involved in their child’s ECEC centre have better social-emotional scores
In Estonia, 80% of parents had a moderate to strong involvement in activities taking place in their child’s ECEC centre, which was higher than in the other two countries participating in the study (69% in England and 65% in the United States). Examples of activities include fetes, concerts/plays, parent’s evenings and parental workshops. Figure 5.16 shows the social-emotional learning outcomes of children whose parents had a strong involvement and those whose parents did not. According to educators, children whose parents had a strong involvement had better social-emotional learning outcomes, after accounting for socio-economic status. However, these differences were not statistically significant for the children’s direct assessment. The positive relationship between parental involvement and children’s social-emotional scores did not significantly differ by gender.
Note: All the differences in this figure are statistically significant.
Every day use of digital devices is associated with lower prosocial behaviour and trust scores than less frequent use
In Estonia, 9% of children never or hardly ever used a desktop or laptop computer, tablet device or smartphone, 13% used them at least once a month, 39% used them at least once a week, and 39% used them every day. The percentage of children who used digital devices every day was similar in England, but around 10% lower than in the United States. Children who used digital devices every day had significantly lower mean prosocial behaviour and trust scores than children who use them at least once a week. There were no statistically significant differences in prosocial behaviour between children who used digital devices every day and those who never used devices. However, children who used digital devices every day had significantly lower levels of trust than those who used them once a month or never used them. There were no differences in the effects of digital device use between boys and girls.
Parents in IELS were asked whether they, or other carers, did educational activities with their children on a digital device (e.g. computer, laptop or smartphone). In Estonia, around 31% of children never or hardly ever did educational activities with their parents on a digital device, 36% did so less than once a week or never, 20% did so one or two days a week, 8% did so between three and four days a week, and 4% did so between five and seven days a week. The percentage of children who did educational activities with their parents on a digital device was lower than in the other two countries participating in the study (the most frequent occurrence in England and the United States was one or two days a week, while in Estonia it was less than once a week). The results did not show a robust association between children who did educational activities on a digital device and those who did not, after controlling for socio-economic status.
copy the linklink copied!Relationship between social-emotional scores and outcomes in other learning domains
Children’s social-emotional skills are associated with emergent literacy and numeracy, after accounting for socio-economic status
Many studies have shown that social-emotional skills are significant predictors of students’ academic performance in areas such as mathematics and reading, after accounting for socio-economic status (Suárez-Álvarez, Fernández-Alonso and Muñiz, 2014[32]; Chamorro-Premuzic and Furnham, 2008[33]). Although previous research has typically assessed students attending primary, secondary and higher education, recent evidence from neuroscience suggests emotion and cognition are interrelated during early infancy development (Bush, Luu and Posner, 2000[2]; Davidson et al., 2002[3]; Posner and Rothbart, 2000[4]).
Figure 5.18 shows the percentage of variation in emergent literacy and numeracy scores explained by social-emotional scores, after accounting for socio-economic status. The first bar presents the percentage of variation in numeracy explained by educators’ indirect assessments of children’s social-emotional skills (prosocial behaviour, trust and non-disruptive behaviour), after accounting for socio-economic background. The second bar shows the association with the direct assessment of children’s social-emotional skills (emotion identification and emotion attribution). The third bar shows the combined effect of the direct and the indirect assessments. While the domains in the second bar were measured using the same assessment method – tablet-based stories and games – the first bar used educator assessments as an independent method. Therefore, the percentages in the first and third bars serve as a proxy of the minimum and maximum variation associated with social-emotional skills, regardless of the assessment method.
Note: SEL = social-emotional learning. The orange bar shows the percentage of variation in each IELS outcome explained by socio-economic status. The green bars show the additional variance explained when social-emotional skills (indirectly assessed, directly assessed or combined direct and indirect) are introduced to a regression model already containing socio-economic status.
The data shows that social-emotional scores were predictive of emergent literacy and numeracy scores in Estonia. Children’s social-emotional scores, together with socio-economic status, explained between 10% and 31% of the variation in emergent literacy scores, and between 5% and 27% of emergent literacy scores, after accounting for socio-economic status. Similarly, children’s social-emotional scores explained between 6% and 26% of the variation in emergent numeracy scores, after accounting for socio-economic status.
Figure 5.19 shows the percentage of variation in mental flexibility, inhibition and working memory scores (self-regulation skills) explained by social-emotional scores, after accounting for socio-economic status. The bars represent the different measures of social-emotional skills based on educator reports, direct assessments or the combined effect of both. Children’s social-emotional scores, together with socio-economic status, explained between 7% and 14% of the variation in working memory scores in Estonia, and between 4% and 11% of emergent working memory scores, after accounting for socio-economic status. Despite sharing the same assessment method, the association between inhibition and empathy skills is weaker. Importantly, educator reports also support the relationship between emotion and cognition in the indirect assessment of social-emotional skills through an independent method. Indeed, educators’ indirect assessment explains a significant amount of variation in self-regulation scores after accounting for socio-economic status.
Note: SEL = social-emotional learning. The orange bar shows the percentage of variation in each IELS outcome explained by socio-economic status. The green bars show the additional variance explained when social-emotional skills (indirectly assessed, directly assessed or combined direct and indirect) are introduced to a regression model already containing socio-economic status.
In short, children with higher prosocial behaviour, trust and non-disruptive behaviour, as rated by their educators, had significantly higher emergent literacy, numeracy, and self-regulation scores. At the same time, the direct assessment showed that children who were more empathetic had significantly higher scores in these domains as well.
copy the linklink copied!Summary and conclusions
The average five-year-old child in Estonia exhibit social-emotional skills at the same or higher levels than their counterparts in England and the United States
When presented with a range of stories and situations, children in Estonia were better able to identify the feelings of the characters than children in England and the United States. Children in Estonia had significantly higher mean emotion identification scores than children in England and the United States.
Children’s ability to recognise emotions is a precursor of the ability to feel empathy towards others, and the emotion attribution scores in IELS reflect children’s ability to recognise their own and other’s emotions. The results show that children in Estonia scored similarly to children in England and the United States in emotion attribution.
According to educators, children in Estonia had significantly higher mean prosocial behaviour than children in the United States and England. However, educators in Estonia rated children significantly more disruptive than children in the United States and England.
Parents in Estonia were more likely to rate children’s empathy skills as more developed than educators (e.g. whether the child is considerate, helpful, caring). Similar findings are observed in England and the United States. This might indicate differences between children’s empathy development in early childhood education and care (ECEC) settings and at home, as well as different expectations of parents and educators. On the other hand, parents in Estonia rated children’s emotional control as less developed than educators. This finding, however, differs from what is observed in England and the United States, where both parents and educators rated children’s emotional control similarly.
Children’s social-emotional skills are related to their emergent literacy, numeracy and self-regulation skills
Children’s social-emotional scores through direct and indirect assessments were excellent predictors of scores in other aspects of the assessment, even after accounting for socio-economic status. In Estonia, children’s social-emotional scores accounted for between 5% and 27% of their emergent literacy scores (compared to between 13% and 33% in England and the United States), between 6% and 26% of their numeracy scores (compared to between 12% and 28% in England and between 7% and 22% in the United States), and between 4% and 11% of their working memory scores (compared to between 7% and 18% in England and between 5% and 22% in the United States), after accounting for socio-economic status.
The development of early skills are interrelated. Cognitive skills are a necessary, but not sufficient, condition to foster early social-emotional learning. For example, children need a minimum level of literacy skills to interact well socially; have rich interactions with peers, friends, and parents; and, ultimately, to open the door to social-emotional learning. However, having high levels of literacy does not always imply high social-emotional skills, and vice versa.
The differences in early social-emotional skills by socio-economic background in Estonia are narrower than in England
Socio-economic background had a significant relationship with children’s social-emotional development in Estonia in emotion identification, prosocial behaviour and trust. However, unlike England, this association was not significant in emotion attribution and disruptive behaviour. The United States showed similar results except in emotion attribution, where the association with the socio-economic background was significant.
Socio-economic background was associated with the frequency with which children engaged with different activities. Parents from a higher socio-economic background were more likely to read to their child from a book, take their child to special activities outside of the home, have frequent back-and-forth conversations with their child, and provide a higher number of children’s books than parents from a lower socio-economic background. These home learning activities are associated with higher scores in social-emotional skills. As in England and the United States, educators in Estonia reported increased parental involvement in the child’s preschool by parents from higher socio-economic backgrounds.
Russian-speaking children have better social-emotional outcomes than Estonian-speaking children
Russian-speaking represent 21% of all children in Estonia. Russian-speaking children had significantly higher outcomes than Estonian-speaking children in emotion identification and emotion attribution, and were rated by their educators as having better prosocial behaviour and trust, after controlling for socio-economic status. As Russian-speaking children generally came from lower socio-economic backgrounds, their scores were even higher after controlling for socio-economic status. According to educators, however, Russian-speaking children were as disruptive as Estonian-speaking children.
While girls typically had better social-emotional scores than boys, the gender gap in Estonia was more prominent than in the United States and England. Furthermore, Russian-speaking children showed more prominent gender differences than Estonian-speaking children.
Differences between Russian-speaking and Estonian-speaking children may relate to different family-based expectations and experiences, although there appeared to be few differences in the activities both groups of parents undertook with their children. Parents of Russian-speaking children in Estonia were as likely to do role-play with their parents, read to their children from a book, and have similar numbers of books at home as parents of Estonian-speaking children. Parents of Russian-speaking children showed a similar level of involvement in the child’s centre as Estonian-speaking children. However, Russian-speaking children were more likely than Estonian-speaking children to use electronic devices as well as go to special activities outside of the home. Nonetheless, Estonian-speaking children were more likely to have back-and-forth conversations with their parents than Russian-speaking children.
Home and family learning environments have a positive relationship with children’s social-emotional scores
After accounting for socio-economic background, the home and family learning environment were a powerful predictor of children’s social-emotional scores. The following factors were positively related to children’s social-emotional scores in Estonia: mother’s tertiary education, having at least one sibling, a high number of children’s books at home, parents who regularly read to their child, going regularly to special activities outside of the home, and parental involvement in the preschool the child attends.
Parents of boys in Estonia were as likely to have back-and-forth conversations with them, read to them from a book, and provide a similar number of books at home as parents of girls. Also, parents of girls showed a similar level of involvement in the child’s centre to parents of boys, according to educators. However, girls in Estonia were more likely than boys to do role-play activities with their parents as well as go to special activities outside of the home.
Children who had experienced social, emotional or behavioural difficulties before the age of five have lower social-emotional skills at age five
Parents in IELS provided information on whether their child had a low birth weight or premature birth, learning difficulties and social, emotional or behavioural difficulties. Additionally, the child’s ECEC centre provided information on whether the child was classified as having special educational needs. The data showed that social, emotional or behavioural difficulties were more highly associated with lower social-emotional outcomes than low birth weight, learning difficulties, or special educational needs – except in the case of trust and empathy skills, where learning difficulties were more highly associated. Disruptive behaviour was particularly associated with social, emotional or behavioural difficulties, as might be expected. When analysed alone, children with a low birth weight had lower emotion identification skills than children born weighing over 2.5 kg. However, these differences disappeared when controlling for children’s learning and social, emotional or behavioural difficulties.
References
[25] Abrahams, L. et al. (2019), “Social-emotional skill assessment in children and adolescents: Advances and challenges in personality, clinical, and educational contexts”, Psychological Assessment, Vol. 31/4, pp. 460-473, https://doi.org/10.1037/pas0000591.
[31] Baldwin, D. and L. Moses (1996), “The ontogeny of social information gathering”, Child Development, Vol. 67/5, p. 1915, https://doi.org/10.2307/1131601.
[20] Baumrind, D. (1968), Manual for the Preschool Behavior Q-Sort, University of California. Institute of Human Development.
[30] Bowlby, J. (1983), Attachment and Loss, Volume 1: Attachment, Second Edition, Basic Books, New York, https://www.abebe.org.br/files/John-Bowlby-Attachment-Second-Edition-Attachment-and-Loss-Series-Vol-1-1983.pdf (accessed on 31 July 2019).
[17] Buchanan, A., E. Flouri and J. Brinke (2002), “Emotional and behavioural problems in childhood and distress in adult life: Risk and protective factors”, Australian & New Zealand Journal of Psychiatry, Vol. 36/4, pp. 521-527, https://doi.org/10.1046/j.1440-1614.2002.01048.x.
[2] Bush, G., P. Luu and M. Posner (2000), “Cognitive and emotional influences in anterior cingulate cortex”, Trends in Cognitive Sciences, Vol. 4/6, pp. 215-222, https://doi.org/10.1016/s1364-6613(00)01483-2 (accessed on 30 July 2019).
[9] Caprara, G. et al. (2000), “Prosocial foundations of children’s academic achievement”, Psychological Science, Vol. 11/4, pp. 302-306, https://doi.org/10.1111/1467-9280.00260.
[22] CASEL (2015), 2015 CASEL Guide: Effective Social and Emotional Learning Programs: Middle and High School Edition, Collaborative for Academic, Social, and Emotional Learning, http://secondaryguide.casel.org/casel-secondary-guide.pdf (accessed on 14 November 2019).
[33] Chamorro-Premuzic, T. and A. Furnham (2008), “Personality, intelligence and approaches to learning as predictors of academic performance”, Personality and Individual Differences, Vol. 44/7, pp. 1596-1603, https://doi.org/10.1016/j.paid.2008.01.003.
[13] Ciarrochi, J., P. Heaven and S. Supavadeeprasit (2008), “The link between emotion identification skills and socio-emotional functioning in early adolescence: A 1-year longitudinal study”, Journal of Adolescence, Vol. 31/5, pp. 565-582, https://doi.org/10.1016/j.adolescence.2007.10.004.
[15] Daniel, E. et al. (2014), “Developmental relations between sympathy, moral emotion attributions, moral reasoning, and social justice values from childhood to early adolescence”, Journal of Adolescence, Vol. 37/7, pp. 1201-1214, https://doi.org/10.1016/j.adolescence.2014.08.009.
[3] Davidson, R. et al. (2002), “Neural and behavioral substrates of mood and mood regulation”, Biological Psychiatry, Vol. 52/6, pp. 478-502, https://doi.org/10.1016/s0006-3223(02)01458-0 (accessed on 30 July 2019).
[24] De Fruyt, F. and B. Wille (2015), “Employability in the 21st century: Complex (interactive) problem solving and other essential skills”, Industrial and Organizational Psychology, Vol. 8/2, pp. 276-281, https://doi.org/10.1017/iop.2015.33.
[10] Duncan, G. et al. (2007), “School Readiness and Later Achievement”, Psychological Association, Vol. 43/6, pp. 1428-1446, https://doi.org/10.1037/[0012-1649.43.6.1428].supp.
[16] Flèche, S., W. Lekfuangfu and A. Clark (2019), “The long-lasting effects of family and childhood on adult wellbeing: Evidence from British cohort data”, Journal of Economic Behavior & Organization, https://doi.org/10.1016/J.JEBO.2018.09.018.
[14] Fontaine, N. et al. (2011), “Predictors and outcomes of joint trajectories of callous–unemotional traits and conduct problems in childhood”, Journal of Abnormal Psychology, Vol. 120/3, pp. 730-742, https://doi.org/10.1037/a0022620.
[5] Garber, J. and K. Dodge (1991), The Development of Emotion Regulation and Dysregulation, Cambridge University Press.
[12] Garner, P. and B. Waajid (2008), “The associations of emotion knowledge and teacher–child relationships to preschool children’s school-related developmental competence”, Journal of Applied Developmental Psychology, Vol. 29/2, pp. 89-100, https://doi.org/10.1016/j.appdev.2007.12.001.
[29] Hinnant, J. and M. O’Brien (2007), “Cognitive and emotional control and perspective taking and their relations to empathy in 5-year-old children”, The Journal of Genetic Psychology, Vol. 168/3, pp. 301-322, https://doi.org/10.3200/gntp.168.3.301-322.
[19] Hogan, A., K. Scott and C. Bauer (1992), “The Adaptive Social Behavior Inventory (Asbi): A new assessment of social competence in high-risk three-year-olds”, Journal of Psychoeducational Assessment, Vol. 10/3, pp. 230-239, https://doi.org/10.1177/073428299201000303.
[26] Kankaraš, M. and J. Suarez-Alvarez (2019), “Assessment framework of the OECD Study on Social and Emotional Skills”, OECD Education Working Papers, No. 207, OECD Publishing, Paris, https://dx.doi.org/10.1787/5007adef-en.
[7] Leerkes, E. et al. (2008), “Emotion and cognition processes in preschool children”, Merrill-Palmer Quarterly, Vol. 54/1, pp. 102-124, https://doi.org/10.1353/mpq.2008.0009.
[4] Posner, M. and M. Rothbart (2000), “Developing mechanisms of self-regulation”, Development and Psychopathology, Vol. 12/3, pp. 427-41, https://doi.org/10.1017/s0954579400003096 (accessed on 30 July 2019).
[11] Rhoades, B. et al. (2011), “Examining the link between preschool social–emotional competence and first grade academic achievement: The role of attention skills”, Early Childhood Research Quarterly, Vol. 26/2, pp. 182-191, https://doi.org/10.1016/j.ecresq.2010.07.003.
[21] Roberts, W., J. Strayer and S. Denham (2014), “Empathy, anger, guilt: Emotions and prosocial behaviour”, Canadian Journal of Behavioural Science / Revue canadienne des sciences du comportement, Vol. 46/4, pp. 465-474, https://doi.org/10.1037/a0035057.
[18] Rutter, M., J. Kim-Cohen and B. Maughan (2006), “Continuities and discontinuities in psychopathology between childhood and adult life”, Journal of Child Psychology and Psychiatry, Vol. 47/3-4, pp. 276-295, https://doi.org/10.1111/j.1469-7610.2006.01614.x.
[8] Schoon, I. et al. (2015), “The impact of early life skills on later outcomes”, http://discovery.ucl.ac.uk/10051902/1/Schoon_2015%20The%20Impact%20of%20Early%20Life%20Skills%20on%20Later%20Outcomes_%20Sept%20fin2015.pdf (accessed on 31 July 2019).
[6] Schroder, H. et al. (2017), “Neural evidence for enhanced attention to mistakes among school-aged children with a growth mindset”, Developmental Cognitive Neuroscience, Vol. 24, pp. 42-50, https://doi.org/10.1016/J.DCN.2017.01.004.
[28] Strayer, J. (1993), “Children’s Concordant Emotions and Cognitions in Response to Observed Emotions”, Child Development, Vol. 64/1, pp. 188-201, https://doi.org/10.1111/j.1467-8624.1993.tb02903.x.
[27] Strayer, J. (1987), “Affective and cognitive perspectives on empathy”, in Eisenberg, N. and J. Strayer (eds.), Empathy and its Development, Cambridge University Press, New York, NY, US.
[32] Suárez-Álvarez, J., R. Fernández-Alonso and J. Muñiz (2014), “Self-concept, motivation, expectations, and socioeconomic level as predictors of academic performance in mathematics”, Learning and Individual Differences, Vol. 30, pp. 118-123, https://doi.org/10.1016/J.LINDIF.2013.10.019.
[1] Thompson, R. (2001), “Development in the first years of life”, The Future of Children, Vol. 11/1, pp. 20-33, https://doi.org/10.2307/1602807 (accessed on 31 July 2019).
[23] Wessberg, R. et al. (2015), “Social and emotional learning: Past, present, and future”, in Durlak, J. et al. (eds.), Handbook of Social and Emotional Learning: Research and Practice, Guilford Press.
Note
← 1. The index of socioeconomic status in IELS comprises household income, parents’ educational attainment and parent occupation.
Metadata, Legal and Rights
https://doi.org/10.1787/15009dbe-en
© OECD 2020
The use of this work, whether digital or print, is governed by the Terms and Conditions to be found at http://www.oecd.org/termsandconditions.
