1 Introduction: the Study of children’s Subjective Well-Being

Today, well-being is an engaging construct used to understand children and young people, although the term itself is often poorly defined and under-theorized (Camfield et al. 2009; McAuley and Rose 2010). In psychology, the study of well-being is a relevant topic in salutogenesis (Antonovsky 1987; Ryff 1989a, 1989b) and positive psychology (Cowen 1991, 1994; Seligman and Csikszentmihalyi 2000). According to the World Health Organization (2013), these research perspectives conceptualize individual health as physical, mental and social well-being and not merely the absence of disease and illnesses. Shifting the focus from malaise to well-being, they provide a positive viewpoint from which to analyze a subject’s relational and social competencies (Delle Fave 2006; Diener et al. 2002; Diener 2009).

By analysis of literature, however, it emerged that there have been relatively few studies of the positive aspects of children’s well-being (Holder 2012). Some of these studies have focused on middle childhood (Guhn et al. 2013), and others have focused on adolescence (Varpu et al. 2016). Furthermore, a significant portion of the literature on well-being relates to measurements (Ben-Arieh 2006; Research Center of UNICEF 2013; Statham and Chase 2010), and another part addresses the economic aspect (Cracolici and Giambona 2010; Di Tommaso 2006; Di Tommaso 2007). Overall, the majority of research has analyzed child psychological well-being using a pathogenic approach (Amerijckx and Humblet 2014), considering well-being and malaise as two sides of the same coin instead of using a multidimensional construct, as affirmed by Headey et al. (1993). According to Anthony and Booth (2017), some theoretical models allow researchers to understand problematic behavior, while a few models focus on child and adolescent well-being.

1.1 Subjective Well-Being and Life Satisfaction

Well-being is a complex concept, and it is one of the most important constructs in the field of positive psychology. In this research area, well-being is considered to be psychological well-being, and more frequently, this term is used interchangeably with subjective well-being (Ryff and Keyes 1995). Recently, Diener et al. (2016) have defined subjective well-being as the set of global assessments people make about their own lives and emotional experiences. Therefore, subjective well-being includes broad appraisals, such as life satisfaction, health and the specific feelings that people experience about particular events that happen during their lifetimes. To satisfy subjective well-being, Tay and Diener (2011) have stated that individuals prioritize the attainment of primary and security needs at the expense of secondary needs. However, fulfilling these different needs has relatively independent effects on their subjective well-being. For example, an individual can benefit from satisfying his psychosocial needs, even though his basic needs are not completely satisfied. According to Diener (2000), subjective well-being is a multidimensional construct that is composed of affective and cognitive dimensions. The cognitive aspect refers to an individual’s degree of satisfaction about his living conditions and quality of life, while the affective aspect concerns the experiences that balance positive and negative emotions (Bradshaw et al. 2011). Positive and negative emotions should be considered as a continuum in which emotions may have different values and weights. Above all, it is important to affirm that the enhancement of positive emotions is strongly linked to an increase in subjective well-being and satisfaction with one’s life (Diener et al. 2014).

In terms of the life satisfaction construct, recent research has supported its key role in the psychological well-being of children and adolescents (Rees and Dinisman 2015). Life satisfaction is defined as an individual’s cognitive appraisal of the quality of childhood life situations (Diener 1984), and it is a crucial mental health component that is directly related to psychological, behavioral and social outcomes. However, life satisfaction can be considered as not only one dimension of the broader construct of subjective well-being (Diener et al. 1997) but also a positive and negative effect balance (Arthaud-Day et al. 2005). When researchers study well-being, a relevant topic is the overall evaluation of life, life satisfaction and commitment within the target group. According to Huebner (2004), global life satisfaction, which is considered to be unidimensional, is measured as a single total score. On the contrary, when global life satisfaction is evaluated through a multidimensional approach, generally several life domains are used. In this regard, the present paper conceptualized life satisfaction as a unidimensional and multidimensional construct to obtain information about the subcomponents of life satisfaction to investigate specific domains, such as satisfaction with family, school, friends, self and living environment, according to Huebner (2004). This approach allows the evaluation of the individuals with regard to various dimensions of life and self.

1.2 The Development in the Ecological Context

Within the research paradigms supporting the ecology of childhood, socio-environmental factors are key determinant of children’s well-being and psychosocial outcomes (Goldfeld et al. 2015; Hertzman 2010). Children’s well-being can be studied through Bronfenbrenner’s ecological system theory (1989; 1999) that focuses on the constant interaction between individuals and their social and physical environment. This theory affirms that children develop in a complex environment and are affected by multiple systems that can be critical to their well-being. The ecological environment includes four subsystems: the microsystem, which refers to the immediate living environments of the child; the mesosystem, which refers to the connections between the child’s microsystems; the ecosystem, which refers to systems with an indirect influence on child’s development; the macrosystem, which refers to the socio-cultural-political system.

From the early years of life, a child’s subjective well-being is closely associated with the relationship that the child establishes in his environment. Among the microsystems of the child’s development, the family is one of the most relevant relational contexts. The family can be a source of well-being and malaise (Barnet et al. 2007; Migliorini et al. 2015b) and it constitutes a pivotal factor in the promotion of a child’s well-being (Dinisman et al. 2017). Relationships within the family are important for a child’s well-being (Emiliani et al. 2010; Fiese 2007; Migliorini et al. 2015a; Migliorini et al. 2016a; Rania et al. 2016) as well as for his overall satisfaction (Andresen et al. 2012). In particular, the quality of parent-child relationship is linked to young people’s well-being (Bradshaw et al. 2007; Navarro et al. 2017; Quilgars et al. 2005), and young people who experience poor parental relationships are at increased risk of problematic behaviors (Olsson and Fritzell 2017). On the contrary, adult connectedness in the family represents a development protective factor (Guhn et al. 2013). Children like good relationships, a supportive and loving family, while disliking conflicts (Andresen et al. 2012; Andresen and Gerarts 2014).

Outside the family, another important microsystem to understand child’s development is the school area (Olsson and Fritzell 2017). Recently, the school context has been identified as a crucial system to facilitate connections between families, children and teachers and as a preventative strategy for child malaise (Cardinali and Migliorini 2013; Schneider et al. 2014). The effect of the school goes beyond that of the family (Sellström and Bremberg 2006) suggesting that deficits and resources of the two contexts produce compensatory effects (White and Warner 2015). Particularly, the relationship with teachers (Guhn et al. 2013) and school climate (Vieno et al. 2005) play a significant role in building a child’s well-being.

Within the ecological model, researchers are also interested in understanding the role of individual characteristics of the child able to influence the level of subjective well-being in the various development contexts. In particular, gender is a child’s variable, as considered by many studies (Cummins 2014): while females result more satisfied than males using multi-item domain-specific scales, the use of general overall subjective measures do not reveal any gender differences (Casas et al. 2013b). Furthermore, studies using items referring to different life domains showed gender differences, with females reporting higher subjective well-being in school and interpersonal relationships (Kaye-Tzadok et al. 2017). In general, research suggests that gender does not have a direct association with subjective well-being, but it is probable that gender is able to condition the effect of other variables, because the process of subjective well-being formation is different for males and females (Chui and Wong 2016).

Regarding the life domain of school, most studies revealed higher satisfaction with school life among girls than among boys (e.g. Casas et al. 2007; Liu et al. 2016). About the possible explanations of this phenomena, research reported boys, compared to girls, as less motivated at schoolwork, spending less time doing homework, having lower expectations and being less enthusiastic about furthering their studies (Cf., Kaye-Tzadok et al. 2017). Chui and Wong (2016) showed that higher academic satisfaction matters for the life satisfaction of girls but not of boys and, on the contrary, that higher academic satisfaction leads to higher happiness for boys but not for girls. Interestingly, the study of Kaye-Tzadok et al. (2017), reporting that satisfaction with school significantly affects boys’ but not girls’ subjective well-being, supports the idea of considering complexity and contingency in the associations with gender and subjective well-being. Also in family life domain, some studies suggested gender variations (Rees 2017), with females reporting higher satisfaction in relationship with parents and in family life (Morales et al. 2013; Strózik et al. 2016). Moreover, always Kaye-Tzadok et al. (2017) found that girl’s family satisfaction have a bigger effect on their subjective well-being compared to males. Nevertheless, in the case of family context, the role of gender is not yet clearly defined and, as suggested by Dinisman et al. (2012), further researches are needed in order to explore gender differences and effects.

2 The Children’s Worlds Study in Italy

The well-being of children is a highly important topic both nationality and internationality (Bradshaw and Richardson 2009; Research Centrer of UNICEF 2013). “Children's Worlds, the International Survey of Children's Well-Being” (ISCWeB) is a worldwide study of subjective well-being that was developed by the International Society for Child Indicators (ISCI). The main purpose of the ISCWeB is to gain knowledge of children’s lives, relationships with family member and friends, daily activities, time use and, in particular, their own perceptions and evaluations of their well-being. The project is based on the idea that to assess whether a particular environment is capable of leading children to develop their potential, one of the key factors to measure is subjective well-being (Dinisman et al. 2015). For this reason, and considering the increasing awareness of the importance of incorporating a child-center perspective regarding children’s subjective well-being (Ben-Arieh et al. 2014), the research directly interacts with children. Children are asked their own opinions and assessments of some variables that determine well-being through a questionnaire that is available in three versions, for children 8, 10 and 12 years of age. The project began in 2009 when a group of researchers, many of whom came from the UNICEF-sponsored International Society for Child Indicators, met in Geneva to discuss these issues to bridge international gaps in the life and psychological well-being of children. In a preliminary phase of the pilot study, countries such as Brazil, England, Germany, Israel, Honduras, Spain and Palestine were involved. Subsequently, after a review of the project, the study was extended to other countries such as South Africa, Romania and Turkey. A third review has also allowed other countries to join, such as Algeria, Canada, Chile, Ethiopia, Nepal, South Korea, Uganda, and the United States. Italy jointed the project in 2013, during the second wave conducted in Liguria (a Northwestern Italian region) by the Department of Education Sciences psychosocial research group. The project received the endorsement of the Italy Regional School Office.

2.1 Italian Context

Italy is a South European country with a population of approximately 60.665.551 inhabitants. The population of the 5- to 9-year-old age group is 2.854.720, which corresponds to 4.7% of the total population (51.5% male). In the last 30 years, Italian society has witnessed an increase in immigration. Considering the entire foreign population of 5.026.153, 6.5% is represented by children 5–9 years of age (51.8% males) (Istat 2016). From the normative point of view, in the last decades in Italy, legislative rules have been established to not only protect children but also to promote their rights and well-being. In 1997, the Italian government issued Law 285, "Provisions for the Promotion of Rights and Opportunities for childhood and adolescence," through the establishment of a specific National Fund for Infancy and Adolescence, intended to respond concretely to the principles outlined in the International Convention of children. That same year, the Parliamentary Commission for Childhood and Adolescence, the National Observatory for childhood, as well as the National Center for Documentation and Analysis for Infancy and Adolescence were established. Furthermore, since 2011, there has been a National Child Custody Authority in Italy. In the last report from the Research Center of UNICEF (2013), considering overall general well-being, health and safety, education, behavior and risks, housing and environmental conditions, Italy was rated 22nd which placed it in the lower half of the rankings, along with three other Southern European countries (i.e., Greece, Portugal and Spain). However, when well-being was assessed based on life satisfaction rated by the children themselves, Italy increased from 22nd to 15th. These data may be related to the assessment of well-being through the analysis of material well-being, health, education, behavior and risks compared to life satisfaction because they do not completely overlap.

2.2 Aims and Hypothesis

Within the theoretical framework presented here, this paper aimed to explore the well-being of 8-year-old children in Italy, according to the ISCWeB project. In Italy, there is no research on the subjective well-being of 8-year-old children and in general, research has produced more evidence about the well-being of children aged 10 and over. In order to fill this gap, the main purpose of the present research was to delineate a descriptive picture of the Italian children’s subjective well-being in this age group, realizing comparisons by gender, nationality of birth (children born abroad vs born in Italy), area of living (urban schools vs rural schools) and type of school (public vs private). The study aimed also to analyze the relations between the life satisfaction measures utilized, to increase knowledge about this aspect. Moreover, in order to enrich the description of Italian children’s subjective well-being, the impact of gender on subjective well-being was considered, exploring more in deep the role of the family and school contexts. Integrating the relative literature, two hypothesis were formulated. First, females compared to males will report higher levels of satisfaction in the family and school life domains. Second, gender will have not a direct association with overall subjective well-being, but it will be able to moderate the effect of school satisfaction on overall subjective well-being.

2.3 Procedure

The study was conducted among primary school pupils aged 8. The members of our research group personally administered the questionnaires in the classrooms, during school hours at the presence of the teachers. This process fully respected the participant’s privacy, anonymity, confidentiality and required approximately one hour for each class group.

In accordance with Italian law, parental consent is obligatory; therefore, we provided all schools with appropriate informed consent sheets. The data collection procedure fully complied with the Research Ethical Code of the Italian Association of Psychology and the ethical recommendations of the Declaration of Helsinki, as well as the American Psychological Association (APA) standards for the treatment of human volunteers.

Before starting the survey, the Italian research team translated the ISCWeB questionnaire from English to Italian using the translation/back translation procedure, and a pilot study was then conducted to assess the measurement tools in the Italian language.

2.4 Sample

The participants’ selection was based on sampling frame consisting in a list of primary schools. A randomized stratified and multi-stage sampling procedure was realized, considering the type (public and private) and location (urban or rural) of the schools. The sampling strategy involved casual selection of class groups within participating schools. The maximum number of groups chosen was decided to be two, in order to obtain the most representative sample. Moreover, one class was extracted in small dimension schools and two classes in large dimension schools, to guarantee to each child an equal probability of extraction.

The study of children’s well-being was carried out among primary school children from third classes. Overall, 50 primary schools joined the survey and 1145 children participated. The average age was 8.21 years (range = 7–10 years; SD = .437 years). Of the total number of participants, 49.85% were male. Considering country of birth, children born in Italy constituted 94% of all respondents and children born abroad the 6%. Finally, 87.9% of the children attended public schools, and 2.9% attended rural schools.

2.5 Instruments

The complete version of the 8-year-old ISCWeB questionnaire was used. The present work considered the following four measurements of overall subjective well-being included in the questionnaire.

  • OLS (Overall Life Satisfaction) is a single-item scale measuring overall life satisfaction (Campbell et al. 1976). It consisted of the following question: “How happy do you feel with your life as a whole?” Children were asked to answer on a bipolar 5-point emoticons scale from 0 = “Most unhappy face” to 4 = “Most happy face”. The scale was transformed into a 0–100 scale.

Usually, the single-item scales are used to represent a global construct and to obtain the subjective perception of multidimensional concepts. The item of the OLS represents the most abstract and least deconstructed level of life satisfaction (Cummins et al. 2003). The measurement of a single item requires the subject to consider all aspects of the phenomena (Casas and Rees 2015; Rania et al. 2015; Youngblut and Casper 1993) and any other deconstructed instrument should display a high contribution when regressed on the OLS (Casas 2017).

  • SLSS-5 (Student’s Life Satisfaction Scale) is a reduced and adapted version of the Huebner scale (1991) (Rees et al. 2010; Casas et al. 2012) that evaluates the satisfaction of one’s life as a whole. Children were asked to respond to five items (e.g., ‘My life is going well’) on a five-point bipolar agree-disagree scale ranging from 0 = “I do not agree” to 4 = “Totally agree”. An index was formed by summing the five items and transforming the sum from 0 to 100 (Cronbach’s alpha .89).

    The SLSS was developed in the US where it has been shown to have good reliability and validity with general samples of young people (Huebner and Hills 2013). During the piloting of this scale within the ISCWeB, it was decided to use only four of the original items, excluding the reversed items, and to add the item ‘the things in my life are excellent’ adapted from the SWLS (Subjective Well-being Life Scale) (Diener et al. 1985) with the aim to improve the scale’s reliability (Casas 2017).

  • BMSLSS (Brief Multidimensional Student’s Life Satisfaction Scale) adapted version (Huebner 1994; Seligson et al. 2003). The scale consists of five item, each of which addresses a singular life domain: ‘Family life’, ‘Friends’, ‘School experience’, ‘Own body’ and the ‘Local area’ (e.g., “How happy do you feel with the area where you live in general?”). Children were asked to answer on a bipolar five-point emoticons scale ranging from 0 = “Most unhappy face” to 4 = “Most happy face”. An index was formed by summing the five items and transforming the sum from 0 to 100 (Cronbach’s alpha .64).

The original scale was developed in the US and has been tested and validated with children and young people aged 8 to 18 (Huebner et al. 2006). The response options in the original version were on a seven-point scale from ‘Terrible’ to ‘Delighted’. In addition to this change, during the piloting within the ISCWeB it was decided to modify the words of the items slightly to increase understanding and comparability among different languages (Casas 2017).

  • PWI4-SC (Personal Well-being Index – School Children). The scale includes four items (e.g., ‘How happy do you feel with all of the things you have?’) from the Personal Well-being Index - School Children (Casas et al. 2013a; Cummins and Lau 2005). Children were asked to answer on a five-point emoticons scale ranging from 0 = “Most unhappy face” to 4 = “Most happy face”. An index was formed by summing the four items and transforming the sum from 0 to 100 (Cronbach’s alpha .56).

The original version of the PWI-SC consisted of seven items and was developed from the adults’ PWI. The PWI-SC was validated in Australia by Tomyn and Cummins (2011). The response options in the original version were on a eleven-point bipolar scale. In addition to this change, during the piloting within the ISCWeB it was decided to adopt unipolar scale, as recommended by the latest manual of the PWI, and to modify the words of the items slightly in order to increase understanding and comparability among different languages (Casas 2017).

3 Data Analysis

Data analysis was performed with SPSS 18 and entailed descriptive analysis of the overall subjective well-being measures, of the satisfaction in the BMSLSS life domains, and of selected items of the ISCWeB questionnaire and regarding specific aspects of family and school contexts. T-tests were performed to realize group comparisons (males vs females, children born in Italy vs children born abroad, urban schools vs rural schools, public schools vs private schools). Correlational and multiple regression analyses were realized to explore the associations between the life domains included in the BMSLSS and the life satisfaction measurements. Finally, a moderated regression analysis was performed to explore the possible interaction effect of gender between satisfaction in life domains (family and school) and overall life satisfaction (OLS and SLSS5).

The sample was weighted by stratum and school dimension, to ensure that the proportion of children in the dataset in each stratum was equivalent to the proportion of children in the corresponding stratum in the population.

4 Findings

The results are structured in three parts. In the first, descriptive and comparison findings concerning overall subjective well-being are illustrated. The second part shows the associations of life satisfaction measures. The third part presents, beside gender comparison findings of selected items regarding family and school contexts, the results of the moderated regression analysis.

4.1 Overall Subjective Well-Being: Descriptive and Group Comparisons

As illustrated by the average values reported in Table 1, a high degree of children’s overall subjective well-being emerges, with the highest total mean on the OLS scale (M = 91.82, SD = 16.98) and the lowest on the PWI4 (M = 71.37, SD = 9.65). Regarding gender, no significant differences emerged. At a descriptive level, children born abroad reported the lowest level of subjective well-being in all four scales. In greater detail, comparing children born abroad with children born in Italy, the latter show the highest level of subjective well-being, both on the SLSS5 (t = −2.356, p < .05) and on the PWI4 (t = −2.419, p < .05). Considering other comparisons, a statistically significant difference emerged on the SLSS5 between children attending urban schools and children attending rural schools, with the latter reporting a higher level of satisfaction with life (t = 3.358, p < .01). Comparing children attending public schools with children attending private schools, the latter reported higher levels of well-being on all four scales, although no significant difference emerged.

Table 1 Total and group mean values of children’s overall subjective well-being
Full size table

Focusing on the satisfaction with life domains of the BMSLSS, the total and group mean values are shown in Table 2. The highest total average value was obtained for ‘Family life’ (M = 3.63, SD = 0.72), followed by ‘Friends’ (M = 3.60, SD = 0.70), ‘Own body’ (M = 3.49, SD = 0.82), ‘School experience’ (M = 3.45, SD = 0.78) and finally ‘Local area’ (M = 3.41, SD = 0.88). In terms of gender, males and females were equally satisfied with ‘Friends’ and roughly satisfied with the ‘Local area’, whereas as hypothesized statistically significant differences emerged for ‘Family life’ and ‘School experience’. Females reported a higher average value in ‘Family life’ (t = −3.586, p < .001) than in ‘School experience’ (t = −.925, p < .01). Children born abroad compared with children born in Italy showed a statistically significant difference for ‘Local area’, for which children born abroad reported the lowest level of satisfaction (t = −2.910, p < .01). There were no statistically significant differences regarding the type of school (public or private) and the location of the school (urban or rural context).

Table 2 Total and group mean values of children’s satisfaction with life domains of BMSLSS
Full size table

4.2 Associations of Life Satisfaction Measures

As shown in Table 3, the correlation between psychometric scales of life satisfaction is moderate, with values ranging from .461 to .590. In addition, the five life domains included in the BMSLSS show significant correlations with the OLS and the SLSS5 (Table 3). With the OLS, the strongest correlation was found for the ‘Own body’ dimension (.344), followed by the other four dimensions, with values ranging from .331 to .288. With the SLSS5, the strongest correlation was observed for ‘Family life’ (.405), followed by the other four dimensions, with values ranging from .402 to .347.

Table 3 Correlations between BMSLSS/BMSLSS domains and satisfaction with life measures
Full size table

Multiple regression analysis (stepwise) confirmed that all life domains included in the BMSLSS are important in determining children’s subjective well-being. Analyzing the standardized coefficients of regression (Beta), all life domains contribute in a significant manner to children’s overall life satisfaction, although differences emerge depending on the scale of life satisfaction considered (Table 4).

Table 4 Regression of BMSLSS life domains on the OLS and the SLSS5
Full size table

The regression analysis for OLS showed that satisfaction with life was most strongly affected by ‘Own body’ (.201), followed in order by ‘Friends’, ‘Family life’, ‘School experience’ and ‘Local area’, with beta values ranging from .170 to .096. Instead, analyzing regression for the SLSS5, ‘Family life’ emerged as the domain that most affected satisfaction with life (.226), followed in order by ‘Own body’, ‘School experience’, ‘Friends’ and ‘Local area’, with beta values ranging from .203 to .143. Overall, all five BMSLSS domains explain 24% of the OLS and 35% of SLSS5.

4.3 Family and School Life Domains: Selected Aspects of Context and the Gender Influence

To deepen differences by gender in ‘Family life’ and ‘School experience’ domains, the present section presents descriptive and comparison analyses of selected items regarding aspects of the family and school contexts included in the ISCWeB questionnaire. Moreover, regression analysis models to evaluate the role of gender as moderator variable between satisfaction in family and school life domains and overall life satisfaction (OLS and SLSS5) are illustrated.

4.3.1 Selected Aspects of Family and School Contexts

The selected items concerning particular aspects of the family and school contexts are shown in Tables 5 and 6.

Table 5 Mean values by gender comparison, in selected aspects of family context
Full size table
Table 6 Mean values by gender comparison, in selected aspects of school context
Full size table

Data analysis of specific aspects of family life (Table 5) revealed that females, compared with males, reported higher average values for most items, with the exception of the evaluation of quality time in family, for which the mean values are similar (Table 3). In greater detail, females reported feeling more listened to by their parents (t = −2.86, p < .01), and they also reported being treated more fairly by their parents (t = −2.00, p < .05). Females were also more satisfied with the people with whom they lived (t = − 2.56, p < .05), and with the other people in the family (t = −2.77, p < .05).

Findings in selected aspects of school life (Table 6) showed that even in this aspect of life, in general, females reported higher mean values compared to males; only in the case of satisfaction with the other children in the class group was the males’ mean value slightly higher. T-tests revealed three statistically significant differences. Females liked going to school more than males (t = −6.018, p < .001), more often reported being treated fairly by their teachers (t = −2.471, p < .05) and were more satisfied with their relationships with teachers (t = −5.056, p < .001).

4.3.2 The Moderation Effect of Gender on Overall Subjective Well-Being

To test the second work hypothesis, and more in general to explore the possible effect of interaction between gender and satisfaction in ‘Family life’/‘School experience’ domains on overall subjective well-being, four moderate regression analysis when gender is the moderator were performed. Table 7 shows two models to study the effect of ‘Family life’ satisfaction on OLS and SLSS5, and two models to study the effect of ‘School experience’ satisfaction on OLS and SLSS5. In the regression models, the satisfaction with ‘Family life’, the satisfaction with ‘School experience’ and participant’s gender were entered in Step 1, whereas the two-way interaction terms were entered in Step 2. Satisfaction with ‘Family life’ and ‘School experience’ were standardized prior to analyses, and the participant’s gender was dummy coded (−1 = males vs 1 = females).

Table 7 The effect of satisfaction with ‘Family life’ and with ‘School experience’ on OLS and SLSS5 when gender is the moderator
Full size table

In Step 1 the satisfaction with ‘Family life’ and the satisfaction with ‘School experience’, were positively associated both with OLS (‘Family life’: R2 = 0.101; F = 63.632, p < .001; ‘School experience’: R2 = 0.109; F = 68.647, p < .001) and with SLSS5 (‘Family life’: R2 = 0.165; F = 94.918, p < .001; ‘School experience’: R2 = 0.162; F = 92.125, p < .001). Results are in line with what emerged in the regression analysis between BMSLSS life domains, OLS and SLSS5 (Cf., par. 4.2.). Further, as predicted, participant’s gender was not related neither with OLS nor with SLSS5. In Step 2 the two-way interaction of ‘Family life X Gender’ (OLS: R2 = 0.102; F = 42.433, p < .001; SLSS5: R2 = 0.168; F = 64.373, p < .001) was not significant neither in the model of OLS nor of SLSS5. Meanwhile the interaction of ‘School experience X Gender’ (OLS: R2 = 0.115; F = 48.447, p < .001; SLSS5: R2 = 0.163; F = 61.697, p < .001) was significant in the model of OLS (b = −0.077, SE = 0.029, p = .007, 95% CI [−0.133, −0.21]). With regard to this significant interaction, simple slop analysis revealed that satisfaction with ‘School experience’ was positively related with OLS for both genders (Males: b = 0.402, SE = 0.40, p < .001, 95% CI [0.322, 0.482]. Females: b = 0.248, SE = 0.40, p < .001, 95% CI [0.169, 0.326]) but the effect was major for males (R2 = 0.154; F = 98.642, p < .001) compared to females (R2 = 0.063; F = 38.493, p < .001) (see Fig. 1).

Fig. 1
figure 1

Interactive effects of satisfaction with ‘School experience’ and participants’ gender on Overall Life Satisfaction (OLS)

Full size image

5 Discussion

Within the context of the ISCWeB conducted in Italy, this paper aimed primarily to delineate a picture of the 8-year-old Italian children’s subjective well-being. In order to do this, overall subjective well-being measures were utilized (OLS, SLSS5, BMSLSS, PWI4), several group comparisons realized (by gender, nationality of birth, area of living and type of school) and the relations between the life satisfaction measures analyzed, including the life domains of BMSLSS. Furthermore, the effect of participants’ gender on subjective well-being was studied, formulating two hypotheses: the expectation of higher levels of subjective well-being for females compared to males in the family and school life domains; the moderation effect of the gender between the satisfaction in the school life domain and the overall subjective well-being.

5.1 Overall Children’s Subjective Well-Being

Overall, analyzing the total mean values of OLS, SLSS5, BMSLSS and PWI4, it emerged that 8-year-old children reported high levels of subjective well-being although, as Strózik et al. (2016) reported, the assessment varies based on the measurement scale utilized, suggesting that the perspective of well-being differs depending on the scale.

As regards the relations between measures, by correlation, it emerged that OLS, SLSS5 and BMSLSS refer to the same construct. To note that the strongest correlation was found between BMSLSS and SLSS5, both multi-item measures although only one is a multi-domain (Cf. Gross-Manos et al. 2015). Focusing on domains of BMSLSS, children reported most satisfaction with the ‘Family life’ domain, according to previous results (Casas et al. 2013b). Moreover, the centrality of family context in determining children’s subjective well-being was confirmed by the regression analysis between life domains of BMSLSS and SLSS5, with the latest more affected by ‘Family life’ dimension, similarly to what reported by Goswami (2012). On the other hand, the regression analysis on OLS revealed the great influence of ‘Own body’ dimension, underlying also the role of satisfaction with self in the affection of children’s subjective well-being (Kaye-Tzadok et al. 2017). Taking all together these first results, two aspects emerge: it is important to consider all satisfaction with life perspectives to assess the level of subjective well-being, because the whole does not simply equate to the sum (Axford 2009; Casas et al. 2013b); the analysis of different aspects may allow researchers to better notice weaknesses or strengths (Amerijckx and Humblet 2014).

Analyzing gender comparisons, no significant differences were observed in the overall subjective well-being scores, showing that 8-year-old Italian males and females are equally satisfied with their lives, similar to reports from other researchers (Savahl et al. 2015). However, more in depth analysis of the ‘Family life’ and ‘School experience’ domains of life significant gender differences were found with females reporting more satisfaction in both domains, in line with the findings of other studies (Bradshaw et al. 2011; Casas et al. 2007; Liu et al. 2016) and confirming the higher satisfaction of females in ‘Family life’ reported by Strózik et al. (2016).

Related to the need and usefulness of adopting a multiple measurement perspective to assess the subjective well-being of children, it is to note that measurement tools may be an important factor in study of the differences in subjective well-being between males and females (Kaye-Tzadok et al. 2017). As suggested by Casas et al. (2013b), it is to understand if males and females interpret questions differently when they are asked about life satisfaction with a single item scale or with differing life domains multi-item scales, and if the domains included in multi-items scales do not cover the aspects of children’s lives important by gender.

Regarding the country of birth, overall, children born abroad showed the lowest ratings of subjective well-being, reporting migration as a vulnerability factor for children’s subjective well-being (Casas and Bello 2012; Dinisman et al. 2012). In particular, significant differences emerged between immigrant children compared with children born in Italy, in the average values of SLSS5 and PWI4. Moreover, when comparing children born abroad with children born in Italy, differences emerged in the satisfaction with ‘Local area’ domain, with children born abroad less satisfied compared to Italian children. In general, these findings could be interpreted inside the theoretical issue, considering migration as a process that requires complex and dynamic adaptations for families (Falicov 2012; Migliorini et al. 2016a). Moreover, looking at the local area, it is also probable that the sense of belonging among migrants playes an important role in the level of satisfaction (Amit and Bar-Lev 2015; Raffaetà and Duff 2013). Beside this aspect, however, it must be considered that in the surveyed area many immigrants reside in suburbs where the physical architecture is less pleasing. In future research, this aspect needs to be verified to see whether a random component subsists in addressing results.

The latest significant difference emerged by group comparisons regards the school location, with children attending rural schools, when compared to those attending urban schools, reporting on the SLSS5 a higher level of subjective well-being. Findings that underscore the need to strengthen the study of the relationships between well-being and environmental factors (Ranisio and Inghilleri 2013), according to the literature show the significance of context to child and adolescent well-being and health (e.g., Burke et al. 2009; Gill 2014; Flouri et al. 2014). For example, as highlighted by the study of Adams and Savahl (2016), the children’s engagement in natural spaces playes a critical role on their subjective well-being, and studies are needed also to understand how benefits can be harnessed to better children’s overall quality of life.

5.2 Family and School Context Life Domains, Gender and Subjective Well-Being

As illustrated in the previous paragraph, gender differences emerged in satisfaction levels of ‘Family life’ and ‘School experience’ confirming, as hypothesized, literature reporting females more satisfied compare to males in the family and school domains. Moreover, by in-depth analysis of gender differences in specific aspects of family and school life contexts, it emerged that females compared to males reported better evaluations of specific aspects of the contexts.

In particular, for the family context, females reported feeling more listened to and being treated more fairly by their parents compared to males, similarly to Morales et al. (2013). In addition, females reported being more satisfied with the people in their families. Findings suggest that family development protective factors, such as adult connectedness (Guhn et al. 2013), may contribute to subjective well-being in the ‘Family life’ domain.

Regarding the school context, it emerged that males are less likely to attend school and more likely to be less satisfied with their relationships with teachers (Cf., Kaye-Tzadok et al. 2017). Furthermore, females more often reported feeling treated fairly by their teachers, according to studies considering the importance of gender differences in contributing to children’s school well-being (Katja et al. 2002; Wang et al. 2016).

Focusing on the moderate regression analysis between satisfaction with ‘Family life’/‘School experience’ and OLS/SLSS5, gender has not a direct association with overall subjective well-being, but it emerged as a variable that significantly affects the relation between ‘School experience’ satisfaction and overall life satisfaction (OLS). Specifically, it emerged that the interactive effect of satisfaction with ‘School experience’ and gender is significant for both males and females, but the effect is major for males. Partly confirming the second work hypothesis, such as the results of Kaye-Tzadok et al. (2017), which found that satisfaction with school significantly affects boys’ but not girls’ subjective well-being.

It is to note that the interactive effect between ‘School experience’ and gender emerged on the OLS but not on the SLSS5. Moreover, no interactive effect emerged between gender and ‘Family life’ satisfaction. About the first aspect, also in this case, it is to consider the differences between single-item and multi-item domain-specific scales because, as mentioned, the single-item requires the subject to consider all aspects of the phenomena without different facets. Instead, regarding the absence of interaction with ‘Family life’ satisfaction, this data contributes to the exploration of differences in gender effect on children’s subjective well-being (Dinisman et al. 2012). As concerns, interesting the findings of the present study differ from that of Kaye-Tzadok et al. (2017), which found that girl’s family satisfaction have a bigger effect on their subjective well-being compared to males. In the interpretation of this difference, such as that regarding the absence of significant effect of satisfaction with school on girls’ subjective well-being, it is to consider that the study or Kaye-Tzadok et al. (2017) employed data from 12-year-old children. Variations in the effect of life domains on subjective well-being due to the transition between the first decade of life and adolescence have to be taken into account, as suggested by gender variations in different life domains (Klocke et al. 2014).

In summary, overall these results confirm that the process of subjective well-being formation is different for males and females (Chui and Wong 2016). However, the nature and cause of associations between gender and subjective well-being are indirect as well as complex (Kaye-Tzadok et al. 2017), and the measurement tools used must be chosen with care in order to capture gender variations (Casas et al. 2013b).

6 Conclusion

The key added value of this study is to have delineated a picture of 8-year-old Italian childrens’ subjective well-being, exploring more in deep the differences by gender in family and school contexts life domains, and the interactive effect of gender in determining overall subjective well-being. The present work presents some limitations that should be considered in order to interpret results and guide future research.

First, family composition and living arrangements of children are variables that several studies on children’s subjective well-being have considered (Cf. Dinisman et al. 2012). This data would have allowed to improve the complex analysis of subjective family well-being, including gender differences.

Second, school variability data would have allowed to study the school context as mediator of the relation between children’s family satisfaction and subjective well-being, as suggested by Olsson and Fritzell (2017). Moreover, children were asked whether they were born in the survey country or abroad. Regarding these data, a possible improvement could be to consider children’s ethnic origins, rather than just their place of birth because migration is a phenomena that must be viewed from a family perspective. In fact, as mentioned above, the family context can define the well-being of the child, and the condition of the migration family may also have an impact on the childrens’ well-being, even if the child was born in the country of immigration.

However, within these limitations, presenting the early findings from the International Survey of Children’s Well-Being in Italy, this paper may be considered as a significant contribution to overcome the gap in the literature concerning the study of the subjective well-being of 8-year-old children. In terms of possible future developments of this work, the contemporary adoption of qualitative methods could be useful to explore gender potential differences in the interpretation of measurement tools and the aspect of life to be included (Casas et al. 2013b). Moreover, the qualitative approach could be useful for cross-cultural adaptations of validated instruments (Glozah 2015; Savahl et al. 2015) and relevant for salient intercultural relationships (Migliorini and Rania 2016a). This issue must be considered in a country with recent immigration, such as Italy.

The subjective evaluations about family context provided by the children themselves can provide guidance for one of the most critical factors of measuring the well-being, that is, the quality of the close relationships in a child’s life (Ben-Arieh et al. 2009). An approach highly relevant for researchers and from a policy-making point of view. Furthermore, the in-depth exploration of the school life domain has identified issues useful for educational policy makers. In fact, some development of these issues could be related to the school experience in Italy, because the early years’ school context has a long tradition of professional development of operators on the relationship with children (Migliorini et al. 2016b). The recommendation is to take them into account as educational models.