Social Relationships , Child Poverty , and Children ’ s Life Satisfaction

Child subjective well-being is determined by various personal, social, and contextual factors. Few studies have found reliable differences in the prediction power of these factors; however, the results vary especially when it comes to sociodemographic factors, such as the effect of child’s socioeconomic background on life satisfaction. This paper examines how poverty and social relationships affect the perceived life satisfaction of Finnish schoolchildren. Drawing on survey data of Finnish schoolchildren, from grades 5, 7, and 9 (n = 1793), linear regression was used to test how life satisfaction would be associated with socio-demographic variables, poverty, and child–parent and peer relationships. The results emphasize the complex nature of the determinants of children’s life satisfaction. The greatest unique contribution for change of life satisfaction was made by the time spent with mother (β(p) = 0.189). Overall, the model showed a good fit (R2 19.9). These findings have important implications for family policies and services that promote good parenting and positive parent–child relationships. Furthermore, this study highlights relational well-being as a key determinant of children’s life satisfaction.


Introduction
The importance of children's subjective well-being is increasingly being recognized, for society as a whole and especially for children (Bradshaw et al. 2011;Casas et al. 2013;Diener et al. 1999;Dinisman and Ben-Arieh 2015).However, much prior research focuses on children's subjective well-being and quality of life from the perspective of adults (experts or parents) (Casas 2011).The results of the relationship between child poverty and subjective well-being of children are mixed depending on the approaches of how to measure child poverty.Some studies using household income or parent-reported material deprivation as the proxy of child poverty generally found no significant or only a weak association between child poverty and subjective well-being of children.(Knies 2011;Rees et al. 2011).Currently, studies have revealed that children's perceptions of material deprivation are related with subjective well-being (Bradshaw et al. 2017;Bárcena-Martín et al. 2017;Main 2018), suggesting that the child-centric method is more effective than the adult-centric method in predicting subjective well-being of children.
In our study, we use a child's perspective in acknowledging the role of children's monetaryand material resources, as well as perceived life satisfaction.Scholars have suggested that these benchmarks capture some of the neglected information on the elementsof children's subjective well-being (Main and Bradshaw 2012;Main 2018).This article examines the life satisfaction of Finnish schoolchildren, based on their own experiences.We are especially interested in how subjective poverty (make ends meet), material deprivation, and children's social relationships affect their life satisfaction.Studies have shown that interactionswith family and friends are important components of children's life satisfaction (Huebner 1991a(Huebner , 1991b)).Research on the connections between children's economic resources and social relationships is surprisingly scarce, but some earlier studies have revealed economic resources can affect children's social relationships as poor children have fewer friends at school and they feel isolated more often (Hjalmarsson and Mood 2015).
We employ a linear regression model to create a view of children's life satisfaction using data gathered directly from children from grades 5, 7, and 9 corresponding to ages 11 to 15. Life satisfaction is a key elementof subjective well-being (Diener et al. 1985;Veenhoven 2012).It is increasingly being recognized in positive psychology as a desiredoutcome for individuals at all stages of life.Life satisfaction of adults has been studied comprehensively (see Diener et al. 1999), but that of children and adolescents has only recently extended (see Huebner 2004).We use a reduced version of Huebner's (1991a) Student's Life Satisfaction Scale (SLSS), which have been tested in earlier studies (e.g., Casas et al. 2013;Rees et al. 2010).By life satisfaction, we mean children's general satisfaction with their lives (Diener et al. 1999;Gilman and Huebner 2003).
From the comparative view, according to a United Nations Children's Fund (UNICEF) report, Finnish schoolchildren score themselves above the average on the life satisfaction ranking compared to children in many other countries.However, Finnish children's life satisfaction declined between 2002 and 2010 (Bradshaw et al. 2013).Compared to many other countries, Finnish children tend to live in families with good material conditions (OECD 2011).However, child poverty rates have increased over the past few years; in 2014, about 11% of all Finnish children lived in poor families (Karvonen and Salmi 2016) and child poverty has been on the rise (OECD 2013).Children's life satisfaction may be affected by the growth of poverty, which can be manifested in social relationships.Firstly, poor children might have fewer friends at school and they easily feel isolated more often (Hjalmarsson and Mood 2015).They also have a highe risk of harassment, they do not participate to school activities and they recive less social support than children from wealthier families (Olsson 2007;Sletten 2010).Second, Finnish children feel pressured by schoolwork, and they do not like school as much as students in other OECD countries (Bradshaw et al. 2013).Third, Finnish schoolchildren experience loneliness.Compared for example to American children, Finnish 10-year-olds, especially boys, are more socially and emotionally lonely (Junttila and Vauras 2009), which consequently may predict lower life satisfaction.Good family relationships and having a large number of friends predict children's happiness (Uusitalo-Malmivaara and Lehto 2013).Clearly, many factors might determine Finnish children's life satisfaction; thus, there was a need for further research.

Prior Research
There has been growing interest in positive psychology to discover in what way and why people experience their lives in positive ways (i.e., subjective well-being).Subjective well-being which encompasses psychological state, for example life satisfaction, and spirituality, that is, personal development and fulfillment (Axford et al. 2014), is usually divided into two components: (1) cognitive evaluations of one's life (satisfaction with specific domains and life in general); and (2) positive and negative emotions or affects at a particular point in time.Subjective perspective to well-being research is very important because it considers person's evaluations of desires, hopes, and plans for the future.Such evaluations concern, for example, body, looks, friends, or feelings of belonging (Axford et al. 2014).In this study, we use the concept of life satisfaction, an overall cognitive appraisal of how good one's life is.(Diener et al. 1999;Gilman and Huebner 2003;Huebner 2004).
Having established that children differ in terms of life satisfaction (see Bradshaw et al. 2013;Casas et al. 2013;Rees et al. 2010), the next question is why.Thus far, the factors of children's life satisfaction are only vaguely understood.As Veenhoven (2012) points out, life satisfaction involves various levels of human functioning including individual behaviour, simple sensory experiences, higher cognition, collective action, the individual's stable characteristics and environment, as well as life chance factors.The literature review summarizes existing knowledge on the determinants of children's life satisfaction.

Demographic Factors
Overall, the demographic traits of children and families usually explain some variation in children's life satisfaction, but the association is unclear.Studies have found out that children's life satisfaction varies by gender and age.Girls have usually lower life satisfaction than boys (Bradshaw et al. 2011;Goswami 2014;Klocke et al. 2014;Rees et al. 2010).Girls usually face greater pressure due to their appearance and weight (Rees et al. 2010;Uusitalo-Malmivaara 2014).In Finland, girls tend to have lower school well-being and are less happy with their confidence, feelings of competence, and safety (Uusitalo-Malmivaara 2014; Uusitalo-Malmivaara and Lehto 2013).All this may explain why girls are less pleased with their lives.However, some studies have not showed significant gender differences in life satisfaction (Huebner 1991a;Ronen et al. 2016).
Age is associated with children's life satisfaction.Life satisfaction decreases when children grow older, especially between childhood and adolescence (Goswami 2014;Klocke et al. 2014).Rees et al. (2010) found that children's happiness with many features of their lives declines as they get older.Older children were less satisfied than younger ones regarding their degree of freedom and autonomy, safety, schoolwork, leisure, appearance, and confidence (Rees et al. 2010).However, the results are inconsistent, as some other studies have not demonstrated differences in children's life satisfaction according to age (Bendayan et al. 2013;Seligson et al. 2003).
Furthermore, children's life satisfaction is associated to the family structure (whether a two-parent, single-parent, or stepparent household).Children living in the household with both biological parents present usually have the highest life satisfaction, while those who live with one parent or stepparents have lower life satisfaction (Bradshaw et al. 2011;Goswami 2014;Klocke et al. 2014;Rees et al. 2010).One theory is that single mothers economic disadvantages, can lead tonegative consequences for children.In addition, psychosocial processes within the family-for example, higher levels of parent-child conflicts-can be a detriment for children in non-intact families (Amato and Keith 1991;Demo and Acock 1996).However, some scholars have revealed that family structure or parents' marital status are only weakly associated with children's life satisfaction, or not at all (Huebner 1991a).The quality of relationships might be more important for children than family structure itself.
Previous studies have also explored how parents' labor market situation affects children's life satisfaction.Usually, children who live in a family with no working adult have lower life satisfaction than children with one or both parents working for pay (Klocke et al. 2014;Rees et al. 2010).Parents' involvement in the labor market is crucial to securing an adequate household income.Employment also increases parents' self-esteem, autonomy, and self-reliance.Thus, parents' employment can increase their children's well-being; it enhances a family's material circumstances, helps to create family routines that support the work ethic, and deliver stability in children's lives (Save the Children 2014).

Social Relationships with Family and Peers
Social relationships are a fundamental human need.For children, parent-child and peer relationships are critical and determine their life satisfaction.Huebner (1991aHuebner ( , 1991b) ) showed that family relationships and friends are two important components of children's life satisfaction (Huebner 1991a(Huebner , 1991b)).Social relationships may explain variation in children's life satisfaction more than sociodemographic factors, or family income and material deprivation (Antaramian et al. 2008;Goswami 2012).
Children's life satisfaction is connected with parental relationships and interactions between other family members (McKeown et al. 2003;Rees et al. 2010).Parental social support and parent-child conflicts, familial stability and mutuality, family functioning, and parental relationships contribute significantly to children's life satisfaction (Gilman and Huebner 2003;Rask et al. 2003).Happiness and satisfaction with family, good parent-child relationships, and family togetherness are the variables most strongly associated with children's sense of contentment, even more than friendships and satisfaction with friends (Gilman and Huebner 2003;Huebner 1991a;Rees et al. 2010;Williams and Anthony 2015).
Relationships outside the family, such as those with peers and friends, are also important to children, especially during late childhood and early adolescence when they move toward autonomy, away from their parents (Giordano 2003).Strong relationships with peers, as well as popularity among them, has positive influence in children's life satisfaction (Gilman and Huebner 2003;Holder and Coleman 2009; Uusitalo-Malmivaara and Lehto 2013; Williams and Anthony 2015).Few friendships predict low life satisfaction, while close relationships with friends enhance it (Uusitalo-Malmivaara and Lehto 2013).Previous studies have generally not predicted life satisfaction based on the number of friends, but friendship quality has predicted life satisfaction more than the influence of the number of friends (Demir and Weitekamp 2007).

Economic Resources
Some quantitative studies have indicated that household poverty and material deprivation are vital features in children's life satisfaction (Bradshaw et al. 2011;Goswami 2014;Lau and Bradshaw 2016;Rees et al. 2010;Main 2018).Children with a greater level of material deprivation or lower household income (reported by parents) have lower life satisfaction than other children (Goswami 2014;Rees et al. 2010).Poverty is connected with children's happiness in several areas in life: material possessions, home, school, and homework (Rees et al. 2011).However, these associations are usually weak (Bradshaw et al. 2011).Children's life satisfaction may be affected indirectly; for example, through the kind of food they eat, the variety of their leisurely activities, the variety of their neighborhoods and schools, or through socioemotional and psychological effects on family relationships (McFall and Garrington 2011).Furthermore, measuring poverty by the child-derived method is stronger in predicting subjective well-being of children (Bradshaw et al. 2017;Lau and Bradshaw 2016;Main and Bradshaw 2012;Rees and Bradshaw 2016).
Some qualitative studies on the consequences of childhood poverty have explored how children experience poverty.Results reveal that the consequences of poverty are social in particular (e.g., Attree 2006;Ridge 2002;Van Der Hoek 2005).It is often not the poverty per se that negatively disturbs children's life satisfaction, but rather the social consequences of economic hardship.Children's involvement in social activities depends on material and economic resources.Low-income children may have less opportunities to engage in social interactions with their peers, and are often excluded from organized leisure activities.Furthermore, their limited ability to match their peers in terms of consumer belongings may lead to social exclusion (Hjalmarsson and Mood 2015;Olsson 2007;Ridge 2002;Sletten 2010).Poor children often have fewer material goods than other children, and therefore might also experience bullying (Ridge 2002;Van Der Hoek 2005).
A lack of economic resources can thus have multiple consequences since they can impact social relationships, in turn possibly determining children's life satisfaction.Poverty and economic hardship can influence family life and children's social relations within the family.Socioeconomic disadvantages are associated with higher levels of marital conflict, parental mental health, and stress (Taylor et al. 2000); this may pressure parents to meet their children's needs during times of economic scarcity, which can lead to frustration and conflicts.Strong family relationships, parental care, and parental support are significant and protective resources in poor children's lives and emotional development (Attree 2004).
Poverty is also related with children's peer relations and friendships.Money permits consumption, which is a way to gain social status and desirability to potential friends.For example, low income children have fewerfriends they meet their friends less and are less popular within their peers (Olsson 2007;Sletten 2010), which may decrease their life satisfaction.Moreover, economic strain can have more unintended effects on social relations by increasing conflict and pressure levels within the family, in turn impacting children's life satisfaction and consequently, friendships.However, Main (2018) suggests that income, material deprivation, and the way children feel that resources are allocated in the family, are all related to the subjective well-being of children.The multifaceted nature of these relationships demonstrates the multi-dimensional nature of material hardhip and its consequences.Household income is an important element; but alone it cannot capture children's roles in weighing their needs and material living conditions.

Aims of the Study
This study examines how subjective poverty, material deprivation and social relationships are associated with schoolchildren's life satisfaction.The study addresses an important gap in knowledge by predicting the associations of social relationships and poverty concerning life satisfaction.Using regression modelling, we predict the impact of each factor using enter method in three blocks.This way we want to control the way variables are included into the regression.The first block contains control variable gender, grade, family structure, and parental employment.The second block adds variables of social relationship and the third block the variables of poverty.
We expect that positive associations will be found between child-parent relationships and life satisfaction: the more time is spent with parents, the higher is life satisfaction of children.We also predict that there is an association between peer relationships and life satisfaction: the more friends children have, the higher their life satisfaction.We posit that there will be negative associations between subjective poverty and life satisfaction: if children have difficulties to make ends meet (lower scores) the lower their life satisfaction is.Finally, we predict an effect between a child-derived indicator of deprivation and life satisfaction: the more deprived children are, the lower their life satisfaction is.

Survey Data
We derived the data from a survey conducted in 2014.The participants were schoolchildren in grades 5, 7, and 9, which corresponds to ages 11, 13, and 15, respectively.The participants completed an online questionnaire during school hours with the items on the questionnaire grouped into logically coherent sections.A total of 1793 child provided data with their parents' consent.They attend 60 schools, mainly in southwestern and southern Finland, where most of the Finnish population lives.There were 938 fifth graders (450 girls or 48% female, and 488 boys or 52% male), 403 seventh graders (211 girls or 52% female, and 192 boys or 48% male), and 452 ninth graders (242 girls or 54% female, and 210 boys or 46% male).The data were diverse both socially and demographically, representing Finnish children quite well compared to official statistics; 24% reported living in single parent homes, and 7% were born abroad or had at least one parent born abroad.According to Statistics Finland (2016) about five percent of the population had a foreign background.Fourteen percent of the respondents reported it was difficult for their family to make ends meet on their family's income.

Life Satisfaction
Life satisfaction, "a global assessment of a person's quality of life according to his [sic] chosen criteria" (Shin and Johnson 1978) forms one dimension of subjective well-being (Huebner 1991a).Five manifest variables were used for the latent variable of life satisfaction.These variables weredapted from the SLSS, which originally was a seven-item self-report scale intended to evaluate students' life satisfaction at a global level (Huebner 1991b).In this study, a reduced five-item version of the SLSS was applied.According to Rees et al. (2010), two of the seven items could be removed from the analysis without changes in psychometric properties, for example scale reliability (Rees et al. 2010).Thus, we asked the children in our study to ponder five life satisfaction items: (1) "My life is going well," (2) "My life is just right," (3) "I have a good life," (4) "I wish I had a different kind of life" (reverse coded), and ( 5) "I have what I want in life."The response scale varied among "totally disagree," "disagree," "neither disagree nor agree," "agree," and "totally agree."We used the measure as a continuous variable by calculating mean scores, so that the minimum score was 1 and the maximum 5.The internal consistency (Cronbach's alpha = 0.869) of the unidimensional factor structure of the SLSS was adequate (confirmatory factor analysis [CFA], extraction method maximum likelihood, direct oblim rotation), and the mean score of the SLSS was 3.99 (SD = 0.88).

Dimensions of Poverty
We measured poverty using two indicators.
The first indicator, subjective poverty, gauges children's self-reported difficulty to make ends meet, and offers important insight.Responses to the question "How difficult/easy it is for your family to get along on your family's income?" had six response options ("1 = very difficult to 6 = very easy").The item was used as a dichotomous variable (0 = poor (very difficult, difficult, somewhat difficult), 1 = not poor (very easy, easy, somewhat easy)).Those who perceived that they experience a high degree to make ends meet represented poor children (14%, n = 227).
The second indicator to measure poverty, material deprivation refers to the lack of necessary components of an acceptable lifestyle, especially those consumption goods and activities that children cannot afford but which are considered typical in a society at a given point in time, irrespective of their preferences with respect to these items.Child-derived material deprivation is obtained from Main and Bradshaw (2012).We added a smartphone to the original 10-item instrument, since smartphones are essential devices for children to access the Internet, communication, and social media.Children can feel deprived if they don't have a smartphone, which 81% of 7-14 year-old Finnish children have (Merikivi and Myllyniemi 2016).Following Main and Bradshaw (2012), we reverse coded and recategorized the item scales resulting in a scale of 0-11, with 0 indicating the highest degree deprivation and 11 of no deprivation (Main and Bradshaw 2012).When we used material deprivation as dichotomous, we employed a cut-off score of 5 (Cronbach's alpha = 0.834).

Parental and Peer Relationships
We used two measures of social relationships in the analysis.
We formed the first one (parental relationships) from eight items that assessed children's relationships with their mothers and fathers; we obtained it from a Swedish study on children's living conditions (Swedish Agency for Youth and Civil Society 2007).These items asked about the time that children spend and bond with their parents (e.g., eating breakfast or dinner, watching tv, doing homework together).Respondents were asked: "How often do you do the following with your parents?" and to choose the best corresponding answer from the Likert scale: ("1 = never to 5 = every day").We used the parental relationship scores as dichotomous variables based on mean scores that varied between 1 and 5; a cut-off score of ≤3 indicates less time spent with parents ("0 = less than weekly to 1 = at least weekly").The reliability coefficient (Cronbach's alpha) in the overall study sample for the relationship with mothers was 0.845, while that of the relationship with fathers was 0.892.
The second indicator, peer relationships (Swedish Agency for Youth and Civil Society 2007), assessed the respondents' number of close friends by asking, "At the moment, do you have a really close friend with whom you can talk about almost all confidential things?".The variable was used as dichotomous ("0 = no close friends, 1 = at least one close friend").

Data on Background Variables
Self-reported gender, grade, family structure, and parent's employment status were used as socio-demographic background variables.Family structure was measured by the question: "With whom you live?" which was reverse coded and recoded into a dichotomous variable (0 = single parent, 1 = both parents).Participants who did not live with either parent (n = 27) were coded as missing.Thirty percent (n = 8) of them attended elementary school and 67% (n = 18) were boys.We gauged the parents' labor market situation by asking "Is your mother/father employed?".

Statistical Analysis
Continuous non-normally distributed outcome variable of life satisfaction was first tested using means, standard deviations, and mean ranks.Mean ranks which were counted over the total data, were used in analyses because the outcome variables SLSS was particularly left skewed in distributions.t-test for independent samples was applied in comparisons between background and outcome variables.Second, all the statistically significant background variables (Close friends, Relationship with mother/father, Subjective poverty, Material deprivation) were subjected to linear regression analysis.The interaction effects between all the background variables and the outcome variable were evaluated which resulted in the exclusion of thevariable of mother's employment status as non-significant from the final regression models.Imputation of the data was not applied to statistical analysis.Statistical analyses were carried out using SPSS 23 for Windows (IBM Corp., Armonk, NY, USA) and p-values below 0.05 were considered statistically significant.We computed Cohen's effect size to test significance that are independent of sample size.

Results
Most of the children were highly satisfied with their lives.The younger the respondent, the higher his/her life satisfaction.Compared to children with higher life satisfaction, those with a lower level satisfaction spend less time with their mothers doing different activities (t(1735) = −12.731,p < 0.001).Likewise, children with lower life satisfaction differ from those with a higher level in terms of time spent with their fathers (t(1726) = −10.338,p < 0.001).Children who are not deprived have higher levels of life satisfaction compared to those who are deprived (t(1614) = −7.126,p < 001).Children who felt their families can make ends meet scored higher on life satisfaction scores (t(1627) = −10.251,p < 001) compared to children whose families cannot.Table 1 shows the descriptive results of the outcome variable and its predictors.In the first block, after introducing the background variables in the model, the reported regression coefficients were significant at a p value of less than 0.001, except for the father's employment status (p < 0.05).The best predictor of life satisfaction is grade; those in elementary school, i.e., the younger students, perceive higher life satisfaction than those in upper comprehensive school (β = −0.210).
In the second block, variables of social relationship, child-parent relationship and peer relationship, were added.As earlier research suggests, child-parent relationship predicts children's life satisfaction well.In this block, we found that the relationship with the mother has the strongest association with life satisfaction (β = 0.217).Peer relations are also found to be positively associated with life satisfaction (β = 0.129).After entering the last block, the variables of subjective poverty and material deprivation, the effects of family structure and father's employment status were found non-significant for life satisfaction.On the other hand, gender effect strengthens along presenting each new block while the effect of grade shows weakening.The reported regression coefficients were significant at a p value of less than 0.001 except for father-child relationship (p < 0.01).(See Table 2) The final model (model 3) shows that when other variables are controlled, mother-child relationship is the strongest predictor of child's life satisfaction (β = 0.189).We also note the association between children's economic resources and life satisfaction, as shown by previous studies.Low subjective poverty indicates high life satisfaction (β = 0.147).Positive associations emerged among children who reported that their families can make ends meet (β = 0.133).The number of friends is also quite strongly associated with life satisfaction (β = 0.112).Overall, elementary school boys, with at least one good friend, having close relationships with parents, especially with mother and a family having an ability to make ends meet are found to be more satisfied with their lives than other children.Parents' labour market situation or family structure has no impact on life satisfaction in the final model.Model 3 explained 20% of the variance in children's life satisfaction.The regression model predicts life satisfaction better in every enter of a new block as the F-test value shows, meaning that the relationship between the model and the outcome variable life satisfaction is statistically significant and model fit improves with each new block.

Discussion
According to our results, Finnish schoolchildren score high in terms of their life satisfaction.Girls have lower life satisfaction than boys, while older children have lower life satisfaction than younger children.Parental labour market status or family structure has no effect on life satisfaction.However, we were not able to study for example the effects work insecurity, working hours or the non-standard working hours that usually affects the temporal structure of family life and may compromise family time and children's life satisfaction.Our most important findings are that social relationships and family's economic situation, i.e., how children feel their family can make their ends meet and material deprivation explain children's perceived life satisfaction.The mother-child relationship plays an important role in children's judgements of their overall life satisfaction.After controlling for socioeconomic background (gender, grade, family structure, and the parents' employment status), this association also holds.This indicates and strengthens that social relationships, positive parental relationships, and peer relationships determine children's life satisfaction.
We found that a children's ability to make ends meet (subjective poverty) was a strong interpreter of higher life satisfaction.Material deprivation predicts life satisfaction when other factors were controlled.This suggests that children's life satisfaction is influenced by material dimensions.As previous research has shown (Olsson 2007) this means that a lack of economic resources limits children's social lives.For example, the lack of right kind of brand shoes or similar clothes than other youngsters wear may cause social pressure.After all, in the modern society consumption and materialism are perceived as central elements in the lives of adolescents (Shim et al. 2011).However, it is important to acknowledge that although a family might have financial troubles, good parental and peer relationships could compensate the effects of poverty.We need more insight into this issue because there might be different trajectories (mediations) through which poverty influences children's life satisfaction (see Main 2018).
Our study confirms previous findings that children's life satisfaction is more strongly connected with the quality of relationships than with family structure.When other variables are controlled, family structure has virtually no impact on a child's life satisfaction.Instead, a good mother-child relationship predicted higher life satisfaction.This shows that schoolchildren's life satisfaction depends on closeness to their mothers which strongly depends on the mother's physical and psychological well-being, in turn influencing the mother's support for her children (McKeown et al. 2003).That is, parenting may be biased by multiple every day hassles or problematic circumstances which decrease child-parent relationship and may lower child's perceived life satisfaction.We also found that children's relationships with their fathers has significant positive associations with their life satisfaction.This conforms prior research finding that shows fathers can make a unique influence to children's happiness and life satisfaction (Amato 1994).The results, thus, suggest that frequently spent time together doing normal every day routines like talking about important things with parents or going outdoors together increase children's life satisfaction.
In addition to parental relationships, peer relationships were associated with life satisfaction: the higher the number of close friends children have, the higher their life satisfaction.This finding is explored in earlier studies too (see Bradshaw et al. 2011;Demir and Weitekamp 2007;Uusitalo-Malmivaara and Lehto 2013).However, in our study, one limitation was that we only measured peer relationships in terms of the number of close friends; an increase in one's number of good friends is associated with life satisfaction to a certain extent but cannot signify a linear increase.
Despite the study's contributions, several limitations should be considered.First, limitations related to sampling representativeness and regional focus need to be discussed.According to the Official Statistics Finland (2016), 70% of the Finnish population lives in metropolitan or suburban areas, while in our study, that figure was 46%.Thus, the generalizability of the results may be limited.However, the participants attend 60 schools, mainly in southwestern and southern parts of Finland where most of the Finnish population is centered.Participating communities characterize current Finnish regional development well, which during the 21st century has been the strongest in the countryside, close to bigger towns.In addition to the impact limitation, another constraint that affects the outcomes concerns the study design, namely, the cross-sectional nature of the data collection.No causal effect assumptions can be made.However, the results might still be widely applicable in the context of similar studies on children's life satisfaction.Although self-report measures produce information about how children conceptualize and experience their lives, some measures, such as

Table 1 .
Socio-economic background variables affecting children's life satisfaction analysed by t-test for independent samples.

Table 2 .
Children's life satisfaction explained by socio-demographic variables, relationship variables and subjective poverty variables *.