Subjective Well-Being and Personal Relationships in Childhood : Comparison of Brazilian and Spanish Children

Abstract This study aims to compare personal relationships satisfaction and subjective well-being of Brazilian and Spanish children and to verify if personal relationships are associated with subjective well-being for the sample. Participants are 6,747 children aged from 11 to 14 years old (M = 12.07, DP = 0.731), 1,020 being Brazilian and the others Spanish. From the total sample 50.7% were girls. Results pointed that the items of personal relationships are grouped into three components related to family, school and friends. Spanish children had higher means than Brazilians' in the items of personal relationships and in the subjective well-being scale. There is a significant association between personal relationships and subjective well-being, and the relationships with family, school and friends are predictors of children well-being in both countries. Family relationships are those with the highest contribution to the prediction of child well-being for both countries, followed by relationships at school and with friends. Personal relationships can be considered a predictor of child well-being with approximately 40% of explained variance for both countries.

Interpersona | An International Journal on Personal Relationships interpersona.psychopen.eu | 1981-6472 Among the life domains, personal relationships are important in the formation and social and emotional development, meeting human needs of intimate connection with others (Myers, 2000).Latin American social studies about personal relationships highlighted its influence on socio-economic aspects, in addition to influencing health through the promotion of healthy habits and the provision of social support (Garcia, Bucher-Maluschke, Pérez-Angarita, & Pereira, 2016).Family relationships and friendships are an important part of social capital in all ages, and influencing on subjective well-being since childhood (Goswami, 2012).The main personal relationships of children are established in the family, friends and school contexts (Kim & Kim, 2013;Strelhow, Calza, Schütz, & Sarriera, 2013).
The family is the first group to which the child belongs, and is where she learns to relate socially with others (Breheny, Stephens, & Spilsbury, 2013;Köster, Cavalcante, Carvalho, Resende, & Kärtner, 2016), developing boundaries, rules and values (Headey, Muffels, & Wagner, 2014;Yeung & Chan, 2014).Recent study shows that the family structure may affect the children's satisfaction with family life and with life as a whole, beside the child subjective well-being, especially in consequence of the relationships developed between the children and their parents (Dinisman, Andresen, Montserrat, Strózik, & Strózik, 2017).From the family, the first relationships with other children begin to settle down, usually with those that are closer, as the friends (Pereira & Garcia, 2012).These early relationships have greater control of parents, but throughout development, the child will have more autonomy in choosing their friends (Vincent, Neal, & Iqbal, 2017).
In the school context, the relationships are established with peers, but also with teachers and they are considered important for the well-being (Santos, Calza, Schütz, & Sarriera, 2013).The satisfaction with other children in the classroom influences satisfaction with student life, and can contribute more than physical aspects of school or even school grades (Casas & González, 2017).The relationship among classmates offers mutual recognition and experience of friendly relations, generates personal identity and conflict resolution (Holfve-Sabel, 2014).
Studies show that relationships provide climate of mutual trust and respect when they are positive (Betts, Rotenberg, & Trueman, 2013;Orpinas & Raczynski, 2016).However, the relationships can be harmful when there are situations of violence and rejection among peers (Ledwell & King, 2015), and there is a declining trend in academic achievement, school involvement (Hammig & Jozkowski, 2013) and also a decrease in wellbeing (Câmara & Tomasi, 2015;Goswami, 2012).A recent study conducted with Brazilian children showed that the involvement with bullying is related to lower levels of well-being, regardless of being in the position of perpetrator or victim, when compared to those not involved in bullying (Alcantara et al., 2017).Relationships with teachers, if positive, make children trust them as reference adults facilitating commitment to the guidelines for their behaviour and conduct (Holfve-Sabel, 2014).
The development contexts can be considered ecological systems, where people are in constant exchange with the environment that surrounds them (Sarriera, 2010).These systems are divided according to their proximity and complexity.Systems where people live and participate are called microsystems and each has specific Subjective Well-Being and Personal Relationships in Childhood 92 physical and material characteristics, where relationships and social roles are developed (Bronfenbrenner, 2002;Bronfenbrenner & Morris, 1998).
A set of microsystems in which the person is located, is a mesosystem, understood as a bigger and community context, formed by the interaction of different microsystems.There are other systems that people do not participate, but that influence in their lives, as the culture where they are inserted and public policies that compose the exosystem.Finally, the macrosystem goes through other systems and covers other aspects such as religion, ideologies, values, health, political and economic systems (Sarriera, 2010).
Each country has different subcultures distributed in their regions, although when compared to other country they can present certain national identity, which most its inhabitants may share.The recent literature on wellbeing presents some cross-cultural studies (Aknin et al., 2013;Kim, Schimmack, & Oishi, 2012;Sarriera et al., 2014), but are mostly carried out with adults, apart from not focusing on personal relationships.It is known that relations with others contribute to mental health, self-esteem, positive behaviour and well-being (Ngai, Cheung, To, Liu, & Song, 2013).Thus, it is important to study personal relationships in Brazil and Spain, countries experiencing different cultures and ways of relating, and to investigate the importance of experienced personal relations for the well-being of children.
In this study we ask the children themselves about their family, school and friends, considering the importance of personal relationships established in these microsystems.From a broader perspective and comparative interest, we study the answers on personal relationships and subjective well-being of children in Brazil and Spain.The study has three objectives: 1) to verify whether the personal relationships' items used in this study are structured in three dimensions: family, school and friends; 2) to compare personal relationships satisfaction (family, friends and schoolmates) and subjective well-being of Brazilian and Spanish children; 3) to verify if personal relationships are associated with subjective well-being of the Brazilian and Spanish children studied.

Method Participants
The

Instruments
Student Life Satisfaction Scale (SLSS, Huebner, 1991).The SLSS is a brief self-report measure (Huebner & Alderman, 1993), consisting of seven items accessing the satisfaction with life as a whole for children and Items 3, 4 and 7 of the scale were not used in this study because they were not collected in both countries, making it impossible to compare data.So, it was used an adapted version of the SLSS, called here SLSS4, composed of Items 1, 2, 5, and 6, answered on a 5-point Likert scale (from 0 = "strongly disagree" to 4 = "strongly agree").The overall Cronbach's alpha for the pooled sample was of .832.
Personal Relationships Items (http://www.isciweb.org).Children answered 18 items about their personal relations, eight on the family, four on school and six on friends.The items were measured using different types of response ranging from 4, 5, or 11 alternatives.Items on frequency have 4 possible answers (0 = never this week to 3 = everyday).The items on satisfaction have 11 possible answers (0 = completely dissatisfied to 10 = completely satisfied), and the other items evaluate how much the participant agrees with the statement with 5 possible answers on Likert scale (0 = strongly disagree to 4 = strongly agree) (see details in Table 1, in the results).

Procedures
In both countries the schools were selected from a list provided by the State Department of Education and contacted in order to obtain authorization to carry out the survey with school children.In Brazil, the participating children signed the Consent term, as well as their parents or guardians and only children who brought the terms signed could participate.In Spain, the authorization for participation is requested to parents early in the school year by the responsible of the participating schools.After authorization, the application was carried out in a previously appointed time with school, lasting about 45 minutes and held collectively in the classroom with the presence of one or two researchers who assisted the children in completing the questionnaires.All ethical procedures we met for both participating countries.

Strategy of Analysis
Initially, descriptive analyses of the data are presented.For all statistical analysis, the scores of the items were transformed into standardized Z scores, as different types of measures were used (as shown in Table 1).To meet the first objective, principal component analysis (PCA) was performed in order to verify whether the personal relationships' items are structured in three dimensions: family, school and friends, and internal consistency analysis was performed using the Cronbach's alpha (Hair et al., 2009).To answer the second objective, a multivariate analysis of variance (MANOVA) was performed to compare the personal relationships satisfaction (family, friends and school) and the well-being of Brazilian and Spanish children.The scores of components obtained from the PCA for family, friends and school, as well as the score obtained by the PCA of the SLSS4 were used as dependent variables of the MANOVA.The country was the independent variable used in the analysis.Finally, to answer the third objective, multiple linear regression analyses were performed to verify whether personal relationships are associated with subjective well-being of Brazilian and Spanish children.
Subjective Well-Being and Personal Relationships in Childhood

Descriptive Analyses
Table 1 shows the means and standard deviations for the 18 items of personal relations as well as for the four of SLSS4 well-being items.It is observed that the averages of Spanish children are higher for all items.The principal component analysis was performed and resulted in three components of personal relationships, which together explained 46.5% of the variance in the Brazilian sample, 47.13% of the variance in the Spanish sample and 47.83% of the variance in the pooled sample.The components obtained are: 1) Family, 2) School and 3) Friends, and factor loadings can be seen in Table 2.Only the loads above .40are presented (Field, 2009;Hair et al., 2009).
It is observed that, in the Brazilian sample, the item with the highest factor loading for the Family component  2).However, the alpha of the School component rises to .728 if the item "Satisfaction with: Other children in your class" is removed.
In the Spanish sample, the item with the highest factor loading for the Family component was "How often do family: Have fun together", for the School component it was also "My teachers treat me fairly" and for the Friends component it was "Satisfaction with your friends".As for Brazilian children, the item "Satisfaction with: All the other people in your family", from the Family component, was the one with the lowest factor loading.For the Friends component it was "Satisfaction with your relationship with people in general" and the item "Satisfaction with: The school you go to" is the one with lower factor loading on the School component, increasing the alpha of this component to .790 if this item is removed.
In the total sample, the highest factor loadings for the three components are for the items "How often do family: Have fun together", "My teachers treat me fairly" and "How often do friends: Have fun together".The item "Satisfaction with: All the other people in your family" is the one with lower factor loading, followed by the item "Satisfaction with your relationship with people in general".The item "Satisfaction with: Other children in your class" presented a factor loading below .40 in the School component for the Spanish children and for the total sample, loading in the Friends component with higher load.Furthermore, the alpha of the School component increases to .732 with the removal of that item.
A principal component analysis was also performed for the SLSS (Table 3) with the four items used in the present study, with Varimax extraction method and assumptions for the analysis were verified (KMO = .797for the pooled sample, Bartlett test of sphericity p < .001).The items were grouped into a single component, confirming the one-dimensionality of the scale and explaining 68.34% of the variance for the total sample.The factor loadings were between .752 and .861.Internal consistency was adequate both for the total sample, as for samples from Spain and Brazil, and all items contribute to the construct.
Subjective Well-Being and Personal Relationships in Childhood

Multivariate Analysis of Variance
Using the data obtained with the PCA results, multivariate analysis of variance (MANOVA) was performed in order to assess whether there are differences between the means of Brazilian and Spanish samples regarding the personal relationships and the well-being.Results indicate that there are significant differences between the means of the components family, friends, school and the SLSS4 considered as a statistical variable for Brazilian and Spanish children (Table 4).When analysing each dependent variable separately (ANOVA), all of them showed significant differences between countries (Table 5).For all, the means of Spanish children were significantly higher than the means of Brazilians', with higher difference between the means of friends.The means of individual personal relationships items can be seen in Table 1.

Multiple Regression Analysis
Three regressions were performed with stepwise method (see Table 6), one for the pooled sample, one for the Brazilian and one for the Spanish sample in order to respond to the third objective of the study: verify whether Subjective Well-Being and Personal Relationships in Childhood 98 Interpersona 2018, Vol.12(1), 91-106 doi:10.5964/ijpr.v12i1.284 personal relationships are predictors of children subjective well-being.The assumptions were verified by residual analysis (Durbin-Watson -Field, 2009), and the three models were adequate (Durbin-Watson for the pooled sample = 1.974, for Brazil = 1.914, and for Spain = 1,989).For regressions, it was considered as the dependent variable the component created from the principal component analysis of the SLSS4.The independent variables used in the regressions were the three components formed by the PCA of the personal relationships: Family, School and Friends presented in the previous analysis.It is observed that all variables are included significantly in the model of the three analyses.For the pooled sample, the explained variance (adjusted R 2 ) of the model is .406,for the Brazilian sample it is .402and .393for the Spanish sample, which indicates that satisfaction with personal relationships explain approximately 40% of well-being, measured by SLSS4.The family variable explains most of the variance for the pooled sample as well as for the two countries, followed by the school and friends variables.

Discussion
This study aimed to verify whether the personal relationships' items used in this study were structured in three the Friends component for the sample of Spanish children, so, it can be that children in the classroom are seen as friends to this sample.
After grouping the items in the three personal relationships components, it was possible to verify the differences between the means of satisfaction with personal relationships and well-being between Brazilian and Spanish children.The results showed that Spanish children have higher means for personal relationships with family, school and friends, and they also have higher means of well-being than Brazilian children.A recent study found highest means of well-being for Spanish children when compared to their peers in Brazil, England, Israel, South Africa, Uganda and the United States (Montserrat, Dinisman, Băļtătescu, Grigoraş, & Casas, 2015).Apart from these countries, the Spanish children also exceeded the average well-being of children from Algeria, Chile, Romania and South Korea (Lee & Yoo, 2015).In the study of Sarriera et al. (2015), Spanish children had the second highest well-being mean, exceeding Uganda, Algeria, South Africa, Brazil, South Korea and England, being Israeli children those with highest well-being.
In the present study, with children aged from 11 to 14 years old, the well-being and personal relationships means were higher for the Spanish sample.Cross-cultural differences are also identified in other studies (Casas, Tiliouine, & Figuer, 2014;Lee & Yoo, 2015;Montserrat et al., 2015;Ximenes, Assis, Pires, & Avanci, 2013).While in Brazil children live in environments marked by social inequalities and high levels of violence and deprivation, in Spain inequalities are smaller and there is more security, allowing greater freedom to children who can live and relate to people in their community.According to the United Nations Development Programme (UNDP, 2014), the Human Development Index (HDI) measures the long-term progress, using as indicators the income, health and education.It can be calculated for municipalities, states and countries, permitting comparing the development of different locations.In 2014, the HDI in Brazil was 0.744 (79 th country), while the Spanish HDI was 0.869 (27 th country).This difference between the development of countries may be reflected in children, influencing child well-being and other variables evaluated in the study.Similar index was used in the study of Lee and Yoo (2015), the authors used the public expenditure ratio with education, child mortality and measures of inequality between countries.Their results points to higher investment in education in Brazil, but in Spain the infant mortality rates and inequalities were lower.
The regressions results indicated that the personal relationships with family, school and friends are predictors of children well-being in both countries, explaining about 40% of the well-being variance.Family relationships are those with the highest contribution to the prediction of child well-being for both countries, corroborating with the literature (Lee & Yoo, 2015).In the family, the first affect relationships are developed (Olszewski-Kubilius, the classroom (Sun, He, & Yang, 2014).Current studies corroborate the idea that personal relationships developed at school are also important to children well-being (Casas, Sarriera, et al., 2014), particularly positive relationship with school friends (Li & Lau, 2012) and the support of teachers associated with the reduced involvement in risk behaviours and increased well-being (Reininger, Pérez, Flores, Chen, & Rahbar, 2012;Tennant et al., 2015).In addition, school experiences may be more related to the students' well-being, preparing them to cope with the demands of life as well as providing knowledge (Wilmes & Andresen, 2015).
With friends, children learn new ways to deal with conflicts and new social skills (Salvas et al., 2014), and they have the opportunity to relate on equal terms and compare themselves to others of the same age and similar skills (Spencer, Bowker, Rubin, Booth-Laforce, & Laursen, 2013).One result that draws attention in this study is that satisfaction with relationships with friends had the lowest contribution to the well-being of children for both countries.One reason may be the fact that 12 years old is the beginning of the phase in which children have more autonomy, and the beginning of the friendship begins to strengthen later during adolescence (McLean & Jennings, 2012).Future studies could check the strength of the relationship between satisfaction with friends and the well-being after 12 years old, which may have greater magnitude in adolescence than in childhood.The stability also influences satisfaction with personal relationships.Children who change school and the place of residence tend to be less satisfied with relationships, especially with friends, since changing sets them apart from people with whom they live, forcing them to establish new friendships (Montserrat et al., 2015).Further studies may verify the satisfaction with friendships relating it to the stability in children's lives.
Personal relationships were investigated in children from 11 to 14 years old, who experience pre-adolescence and early adolescence.Being a cross-sectional study can be considered one of the limitations of this study, since longitudinal studies could check personal relationships and the children's well-being in different stages of development, bringing new findings about these relationships.Similarly, qualitative studies investigating the children's perceptions of their personal relationships and their well-being could clarify aspects that are not covered by quantitative studies.
Children of a single Brazilian state composed the sample of this study, which can be considered a limitation, as it is not a representative sample of the country.Future research with samples of children from different regions of Brazil could make the comparison between children of the two countries more robust, and provide data to assist in the construction of public policies to improve the quality of life and increase Spanish and Brazilian children well-being.
adolescents between 8 and 18 years.Originally there are seven items: 1.My life is going well, 2. My life is just right, 3. I would like to change many things in my life, 4. I wish I had a different kind of life, 5.I have a good life, 6.I have what I want in life, 7. My life is better than most kids.Sarriera, Casas, Santos et al.

a 0 =
strongly disagree to 4 = strongly agree.b 0 = never this week to 3 = everyday.c 0 = completely dissatisfied to 10 = completely satisfied.Principal Component Analysis (PCA)First, the Z values were used as variables in performing the principal component analysis (PCA).The PCA was performed separately for each country, and also with the pooled sample.The analyses of principal components were performed with Varimax extraction method and assumptions for the analysis were verified (KMO = .862for the Spanish sample, .837for the Brazilian sample and .871for the pooled sample, Bartlett test of sphericity p < .001).Sarriera, Casas, Santos et al.
was "My parents listen to me and take what I say into account", for the School component it was "My teachers treat me fairly" and for the Friends component it was "How often do friends: Have fun together".The item "Satisfaction with: All the other people in your family" from the Family component was the one with the lowest factor loading, for the School component it was "Satisfaction with: Other children in your class", and for the Friends component it was "I have enough friends" for the Brazilian children sample.Cronbach's alphas were adequate for the three components of personal relationships in the Brazilian sample (Table dimensions: family, school and friends; to compare personal relationships satisfaction and subjective well-being of Brazilian and Spanish children; and to verify if personal relationships are associated with subjective wellbeing of the children studied.It was found that the items of personal relationships were grouped in three components: Family, School and Friends.The item "Satisfaction with: Other children in your class" loaded in Sarriera, Casas, Santos et al. 99 Interpersona 2018, Vol.12(1), 91-106 doi:10.5964/ijpr.v12i1.284 Lee, & Thomson, 2014).It is a context where children learn to respect and establish positive relationships of love and respect to others, providing security and tranquillity(McAuley, McKeown, & Merriman, 2012;Scrimgeour, Davis, & Buss, 2016).The family is also primarily responsible for meeting the basic or survival needs, closely related to the assessment of life and affects, that are well-being components(Tay & Diener, 2011).Satisfaction with the relationship at school was also significantly associated with the well-being of children in both countries, pointing to the importance of the relationship between children and teachers and peers at school.Studies show that students' progress depends on the expectations that teachers lay on them and how they recognize their achievements, which directly impacts children self-esteem.The construction of knowledge and the personality development depend on the established relationship with teachers and the other children in Subjective Well-Being and Personal Relationships in Childhood 100 Interpersona 2018, Vol.12(1), 91-106 doi:10.5964/ijpr.v12i1.284

Table 1
Personal Relationships and Well-Being Descriptive Data

Table 2
Personal Relationships Principal Component Analysis -Factor Loadings

Table 4
Multivariate Analysis of Variance Considering Countries for Family, Friends, School and SLSS4 a Dependent variables: Family, friends, school and SLSS4 Component Scores.

Table 5
Analysis of Variance Considering Countries forFamily, Friends, School and SLSS4

Table 6
Multiple Regression Analysis a Dependent variable: SLSS4 Component Score.