Subjective well-being of children in residential care centers : Comparison between children in institutional care and children living with their families

Most of studies with children in residential care centers in Brazil emphasize the problems associated with being in care but few investigate the well-being of these children. Thus, the aim of this study was to investigate and compare the subjective well-being of children in residential care centers versus those from the general population living with their families. Subjects were two hundred and eighteen children, ages 8 to 12 (M = 10.06; SD = 1.40). Half of them are under residential care and half are living with their families. Discriminant analysis was performed using as independent variables the seven PWI-SC items, the eight GDSI domains and the OLS single-item scale, and as dependent variable the group in which the children belong (foster care/families). Results indicate that all items significantly discriminated towards children living with their families. We argue that differences may be related to transitions and life experiences prior to institutionalization. Children with fewer changes manage to maintain their social bonds and find greater consistency in care, which seems to influence their sense of well-being. We conclude that the promotion of well-being must aim at forms of care involving greater stability.


Introduction
Studies with children in residential care in Brazil have been the focus of attention in recent years.However, most of the literature emphasizes the problems associated with residential care and few investigate subjective well-being of these children.The subjective well-being is a psychosocial component of quality of life, and might be understood as a result of the evaluation that people make of their overall lives, taking certain areas into account (Casas, 2011;Diener, 2012).These assessments have two main components, one cognitive and another affective, that are inter-related and express beliefs and desires that individuals have over their lives.Thus, subjective well-being reflects the extent to which individuals believe in achievement and desirability of their lives (Diener, 2012).
The contexts in which children are embedded influence on their well-being and related aspects.Material goods (housing structure, internet, quality clothes to go to school) and cultural objects (books, newspapers) to which they have access, also the relationship they establish with adults and the stability experienced seems to influence the children's well-being (Casas & Bello, 2012;Montserrat, Casas, Malo & Bertran, 2011).Sarriera (2010) points out that the context in which individuals are placed and where relationships are established influence their development.The family is an important context in children's lives.It is their primary socialization instance, considering it facilitates the acquisition of skills and behaviors (Oliveira, Siqueira, Dell'Aglio, & Lopes, 2008).
According to Brazilian law, family is in charge of care and ensuring preservation of the rights of children and adolescents (Estatuto da Criança e do Adolescente [ECA], 1990).However, the dynamics of family functioning may promote both protective and susceptible to risk environments.In families where there are situations of violence, sexual abuse and neglect there is difficult to provide support and protection for children, risking the development and well-being.In such cases, protective measures, such as residential care, become necessary (ECA, 1990;Siqueira & Dell'Aglio, 2007).
The conception that family relations and residential care relations can influence the children's well-being is supported by the premise that individuals can only be understood in their context.Meaning is assigned to social and collaborative environments in context, through reciprocal relationships (Kelly, 2006).Meanings shared by children in residential care may influence their wellbeing differently than those of children living with their families.
Existing data on children's well-being points out that the context prior to foster care exposes children to social inequality, poverty or fewer educational opportunities (Gonzalez, Abaid & Dell'Aglio, 2011).An investigation conducted in Spain, aimed to find the main differences between adolescents living in residential care, adolescents living with one parent and those with intact families.The results indicated significant differences between the well-being of adolescents living with their parents and those in single parent configuration or residing in care, with the highest mean of well-being in the first group (Dinisman, Montserrat, & Casas, 2012).
In the Brazilian context, comparisons between the life satisfaction of children at reception and children living with their families have already been made.Dell'Aglio and Siqueira (2010) investigated life satisfaction among children and adolescents in residential care and children living with their families.The results indicated that children in care had higher average scores of stressful life events and social support.Moreover, the same study points out that fewer life events stressors, greater social support network and less conflict in the family appear as predictors of life satisfaction (Dell'Aglio & Siqueira, 2010).
In a research regarding the subjective well-being of children who live with their family, compared to children living in residential care, Poletto and Koller (2011) found that children in care experience had more negative affections.However, institutionalized children did not differ from those who were living with their families in life satisfaction and positive affect.
For purposes of this study, residential care was assessed.This service shall have similar structure of a residence and must be inserted in communities.It's main goal is to offer personalized attention and to encourage family and community life (Grupo de Trabalho Nacional Pró-Convivência Familiar e Comunitária, 2009) as predicted at the Brazilian Law n. 12010 (Lei n. 12.010, 2009).According to the Institute of Applied Economic Research (Instituto de Pesquisa Econômica Aplicada [IPEA], 2004) most fostered children and adolescents in Brazil are male and the ages range from seven to 15 years old, with over 30% fostered for a time ranging between two and five years.In 2009, Lei nº 12010 has determined as 24 months maximum stay in residential care.
In this sense, authors such as Wade, Biehal, Farrelly, and Sinclair (2011) suggest that one of the problems of protection systems is precisely this temporal and restrictive view based on the dichotomy of children protection system.In this way children will either return to their families of origin, or go to the adoption, where a very important part of these children will not have either of these opportunities.
The main reasons for the residential care are related to problems of parents or guardians, such as chemical dependency, neglect, abandonment, violence, sexual abuse, and imprisonment.There are other issues that lead to institutionalization, such as lack of material resources, orphanhood, sexual exploitation and labor exploitation (Conselho Nacional do Ministério Público, 2013).Carbone, Sawyer, Searle and Robinson (2007) found that children in residential care exhibit losses in development when compared to children who reside with their families.The foster children have more psychiatric disorders and are more difficulty in school demands and to meet and socialize.In a study on the development of children and adolescents in care, Chaves, Lima, Mendonça, Custódio and Matias (2013) found that although these children present development consistent with the criteria adopted by the Ministry of Health also exhibit important social and psychological delays, among which stands out the language.Harden (2004) stresses the importance of stability for the healthy development of children, especially those in care.For this author stability is a facilitating process of healthy development, especially when considering that children in care were exposed to many risk situations.To Oliveira and Milnitsky-Sapiro (2007) instability in residential care (in the exchange of educators, or transfer between institutions) makes bonding more difficult for these children.Considering the importance of understanding the situation of children in care and the extent to which the institution of care itself is a context of healthy development, the aim of this study is to investigate the subjective well-being of children in residential care and compare it with the well-being of children residing with their families that probably have not suffered neglect, abandonment or abuse.

Participants
Participants were two groups of 109 children (total sample of 218).The first group is composed of 109 children from 8 to 12 years old (M = 10.17;SD = 1.42) that live in residential care.For this group, children were contacted through their institutions of care of the city of Porto Alegre.Of those, 62.4% are boys and 37.6% are girls, 41 are aged between 8 and 9 years old and 68 between 10 and 12 years old.All children living in residential care who were within the studied age group were invited to participate and all have accepted.
Participants in the second group are part of a previous study conducted by the Research Group on Community Psychology (GPPC), in collaboration with International Society for Child Indicators (ISCI).The group consists of 109 children randomly selected from a sample of 914 children aged 8 to 12 years (M = 10.07;SD = 1.38), students of public schools from Porto Alegre.Of those, 42.2% are boys and 57.8% girls.From this sample, 41 of 8 and 9 years old and 68 children from 10 to 12 years old were randomly selected so they could be compared with children in the first group.

Instruments Well-being in Childhood Questionnaire
A questionnaire on well-being in childhood, developed in partnership with the ISCI was applied.The survey questions are designed to explore perceptions and evaluations of well-being and its aspects.All instruments were translated from their original languages (English and Spanish) and submitted to back translation.Two pilot applications prior to final version of the questionnaire were conducted in Portuguese.Considering the small number of studies on the wellbeing of children that are answered by themselves, it was deemed necessary to check whether the form of formulating the questions were being understood.Through this application, it was decided that children of 8 and 9 year old would respond to 5-points-items, using emoticons that represent expressions ranging from dissatisfaction to satisfaction.For children aged 10 to 12 years, scales ranging 0 (extremely unsatisfied) to 10 (extremely satisfied) were used.Later, the 5-point scale answers were converted to a 10-point scale for analyzes purposes.
Next, the pilot of the second version of the questionnaire application aimed to verify the semantic understanding of children around the surveyed items and its cultural appropriateness.With due reformulations, the final version of the questionnaire contains sociodemographic variables such as age, sex, city, nationality, and scales that assess the well-being and associated aspects.The scales used to assess the well-being of children in residential care and of those living with their families are described below.

Personal Wellbeing Index -School Children (PWI-SC)
Personal Wellbeing Index was developed by Cummins, Eckersley, Van Pallant, Vugt, and Misajon (2003) and initially directed to evaluate the well-being of adult populations.Originally, the instrument uses seven satisfaction items representing different areas which include the concept of well-being.The version showed high internal reliability, with Cronbach's alpha between .70 and .85.The items are answered from a question of satisfaction with life in general.A PWI version has already been adapted to Brazilian adolescents by Casas et al. (2012) and a Cronbach's alpha of .78 was found.
For purposes of this study, an adapted version of PWI, the PWI-SC, was applied.This version's language is adapted and simplified to school children (Cummins & Lau, 2005).For this sample, Cronbach's alpha is .64.This scale's items concerns satisfaction with all things they have, their health, things they want to be good at, their relationship with people in general, how safe they feel, about doing things away from home and about what may happened to them, later in life.

Overall Life Satisfaction Scale (OLS)
The OLS is a single item that verifies overall satisfaction with life.Campbel, Converse and Rodgers (1976) pointed out the importance of using a single item in the assessment of subjective well-being.This question was asked in different ways according to age.Younger children responded to the question "How happy do you feel with your overall life?" and older children responded to "How satisfied you feel with your overall life?"

General Domain Satisfaction Index (GDSI)
To measure life satisfaction in different domains, Casas and Bello (2012) proposed the General Domain Satisfaction Index.It is originally composed of 29 items that evaluate well-being in eight domains (family and home, material aspects, interpersonal relationships, neighborhood, health, time management), in an 11-point scale ranging from 0 (completely unsatisfied) to 10 (completely satisfied).
The index calculation is performed according to the mean of each domain and, subsequently, with the calculation of an overall mean for all domains.In this study, the totality of items in the index was not used, due to comprehension problems detected in pilot applications).Items regarding local police, the use of time, opportunities in life weren't applied and items "satisfaction with yourself" and "satisfaction with your selfconfidence" were not answered for younger children.Two items regarding family satisfaction had language adapted to foster children's reality.The Cronbach's alpha for the sample of 8 to 9 years old is .88and for the sample of 10 to 12 years .87.

Issues specific to children in residential care
Period in residential care: time, in months, they've been fostered.Some of the children were unable to provide this information and then this information was obtained from legal guardians.
Contact with parents: children were asked whether they had contact with parents and about the frequency of such contact (from never to always), by modality (speaking on the phone, visits to the institution or other prearranged locations or through periods in their fathers', mothers' or adult's homes).
Siblings in care: regarding the quantity of brothers and sisters in protection system.
Institution satisfaction: measured through single item "Since for now you can't live with your family, how happy are you for living in residential care?" with answers ranging from 0 (not at all) to 4 (very happy).
GDSI adaptation: items regarding satisfaction with "the house where you live" and "your family life" were replaced by "the place where you live" and "your life at the place you live", respectively.

Procedures
This research was previously submitted to the Ethics Committee in Research of the Institute of Psychology of Federal University of Rio Grande do Sul and registered with the Platform Brazil under number 00674612.6.0000.5334.After approval Consent Terms were requested from institutions for the research.

Application in residential care
An authorization to contact the institutions of residential care was requested to a City Hall department, and, after approval, a meeting was held at the department to present the project to the team responsible for residential care.
The applications were scheduled in advance by contacting the coordinators of homes.Thus, application did not coincide with school hours or other routine or essential activity for the child.Every child from 8 to 12 years old was invited to participate.The research was explained to the children who have been duly authorized by their legal guardians, through the signing of the Consent Terms.
The implementation of the instruments was given individually, with the assistance of a trained researcher, as recommended by their legal guardian, considering their reading difficulties.The instrument was administered in children's homes and the time of application was approximately 30 minutes.

Application in schools
This research was presented in public schools of Porto Alegre and metropolitan area, which were selected by convenience.Schools that have accepted the research had their students, from 3rd to 7th grade, invited to participate.Students of all classes of these series were asked to answer the questionnaire.
Researchers visited classrooms and distributed the Informed Consent Terms that should be taken home and signed by parents, and themselves.Children who brought the Informed Consent Terms signed and handed them over to researchers, completed the questionnaire.The application was collective and occurred during the school hours, in their own classrooms, or in rooms indicated by school and lasted approximately 40 minutes.Two trained investigators applied the questionnaires.In the beginning of application, it was emphasized that the children were free to participate in the study, and could quit filling out the questionnaire at any time, with no any consequence in relation to that choice.

Analysis Procedure
Descriptive analyzes for the characterization of both groups (children in residential care and children living with their families) were performed.In order to evaluate differences between well-being variables Analyses of Variance (ANOVA) were taken.The first one, performed only with children at residential care, was a Factorial ANOVA to verify differences in well-being considering the independent variables: Period in residential care (4 levels), age (2 levels), sex (2 levels), number of siblings in care (4 levels), contact with the father (3 levels) and contact with the mother (3 levels).The others were three One-Way ANOVA that aimed to investigate the differences between the means of well-being of the two groups of children (residential care or living with families), having as dependent variable the means of PWI-SC and GDSI, and the single item OLS, and independent variable group to which it belongs.
Then, the data were subjected to Discriminant Analysis (DA) in order to specifically check which items of the well-being scales discriminate the two groups.The goal of DA is to differentiate investigated groups by several variables, where the differences are maximized.Independent variables were the items of the PWI-SC, the areas of GDSI scale and the single-item OLS.As the dependent variable, the group to which it belongs (two groups: children who live with their families and children in residential care).

Children in residential care
Contact with parents: 35.8% of children have seen their father, and 66.8% have seen their mother.In addition, 30.3% of fathers and 10.1% of mothers of children have died or are unknown and 74.4% of children live with other adults they consider relevant.Adults considered relevant by the children were uncles and aunts (25.7%), grandparents (22.9%), godmothers (3.7), siblings (14.7%), stepmother / stepfather (3.7%), teacher (1.8%) and cousin (0.9%).It is noteworthy that 24.2% of children whose father are dead or unknown have no contact with any other significant adult in their lives.However, compared to children whose mothers are dead or unknown, all have contact with an important adult in their lives.
The contact that children have with their father, mother or other adults is conducted via phone, visits to the institution or other prearranged location or through periods in their fathers', mothers' or adults' houses.The frequency of contact was measured at an interval of 4 points that ranged from 0 (never) to 3 (always).The highest means of contact with parents (M = 0.40, SD = 0.86) and other adults (M = 0 90, SD = 1.06) are through periods at home, and mothers through visits (M = 0.93, SD = 1.00).

Children living with their families
Children answered a question with regard to the amount of homes that spend most of the time.The answers were that 82.8% of children live in a home and 17.2% in more than a home.In the house, or the house in which they spend most of the time, 95.3% of children live with their mother, 59.4% live with their father and 62.5% live with their siblings.Regarding the partners of parents, 20.3% live with their stepfather and 3.1% with the stepmother.Regarding members on the extended family with which the children live 28.1% live with their grandmother and 4.7% with grandfather.In addition, 86.8% of children currently reside with the same caretakers as in the last year, 65.7% in the same house and 74.2% live in the same neighborhood as they did in the last year and 66.2% have not changed school in the last year.

Differences between children in residential care by the period in residential care, sex, age, siblings in care, contact with father and mother
In Table 1  Factorial Analysis of Variance (ANOVA) was performed to check the differences between children's well-being means, according to period in residential care, siblings in care, sex, age, contact with father and with mother.There were no statistically significant differences (p > .05for all variables) between groups for the mean of the three measures of well-being (PWI-SC, GDSI, OLS).However, the girls showed lower means than the boys for the three measures.In PWI-SC, children with the lowest means are those who do not live with any sibling in care.In GDSI, children with the lowest means are those who have no contact with the mother, while the lowest mean for OLS was for children who have no contact with father.

Differences between means of well-being of children is residential care and children living with their families
Table 2 presents the means of the items and the scales of well-being for the group to which it belongs.Three One-Way Analysis of Variance were performed to investigate the differences between the well-being of children residing with their families and well-being of children in residential care, according to their ages (Table 3).The means of the well-being scales PWI-SC, GDSI and single-item OLS were used as dependent variables.The group (children in residential care and children living with their families) was the independent variable.Note: Independent Variable = Group (children in residential care/living with family).

Discriminant profiles of children who reside with their families and in residential care
In order to verify to what extent the observed differences between the groups show a discriminating profile of variables, Discriminant Analysis was performed.Independent variables were the items of the PWI-SC (seven items), the mean of the GDSI domains (eight domains) and the single item OLS, and dependent variable was the group to which belongs (two groups: children in residential care and living with family).
The number of discriminant functions is determined from the number of groups minus one.Thus, a single function was obtained since two groups are considered.Being unique, the function obtained in the analysis between groups explains 100% of the variability among them.Next the variables that are significant and more relevant in discriminative ability between groups of children who reside with their families and in residential care in order of absolute size of correlation in the structural matrix are presented (Table 5).
Table 5.The group of children who live with their families presented the centroid value of .682,while the group of children who are in residential care showed the value of-.674.Through these values we find that the groups are widely separated by the profile obtained, and the sign of the correlations indicates favorable direction toward one group or another.Thus, it is observed that all variables showed significant positive signal discriminating in favor of children who reside with their families.

Structural matrix of the canonical discriminant function: Correlations between discriminating variables and standardized discriminant function
It is possible to observe that the seven items of PWI-SC, the eight domains of GDSI, and the OLS significantly discriminate children residing with their families of those who are in residential care, in favor of the first, since the values of the structural matrix are positive.However, the PWI-SC item "satisfaction with doing things away from home" and the GDSI domain "Satisfaction with the neighborhood" both have values below .20,which was considered the cutoff point for this study, and therefore did not contribute to discriminate the groups.

Discussion
This study aimed to investigate the subjective well-being of institutionalized children and compare it with that of children living with their families.The children in care belong to a portion of the population which is usually neglected in studies that aim to investigate positive aspects of their lives and those of their peers.
The institutions participating in the research are still in adaptation Law n. 12010 ( 2009), with respect to the maximum period in care.It is found that almost 12% of the sample of institutionalized children is in residential care for more than 24 months, this being the maximum time allotted legally.However, in a previous study in similar context, even greater periods of care were found, reaching a maximum time of 6 and half years of institutionalization (Gonzalez et al, 2011).
It appears, however, that even if the children are still upheld by higher time than they should according to the law, the time in care did not differentiated the well-being variables significantly (PWI-SC, GDSI and OLS).This aspect of the time children remain in care has been approached in recent investigations (Wade et al., 2011, Del Valle, Bravo, Álvarez, & Fernanz, 2008) claiming the importance of stability other than the exact amount of time spent in protection system.Children should be protected for the time they need and if they can't return to live with their families and can't be accepted by others, they should remain -in good condition -in the system.Studies are not showing that children that are in care for the longest are worse, on the contrary, research demonstrates that children who have stability, have better well-being (Montserrat, 2012).
Children who are in residential care presented lower means than children living with their families in the three well-being measures applied (PWI-SC, GDSI and OLS).Similar results using other methodological strategies were found by Carbone et al. (2007) in research with children and adolescents in residential care.The researchers found that children and adolescents in residential care are in disadvantage when compared to children who reside with their families, considering their lowest levels of well-being.The differences between living in residential care, at foster family, at foster home or with parents were not evaluated in this study.
It is understood that the perception of well-being of children in residential care is also influenced by their previous life experiences, considering that the main reasons that lead to their institutionalization are abandonment, neglect and maltreatment (IPEA, 2004).The removal of the child from the family environment is a protective factor for issues such as poverty, lack of social support, mental illness --with or without substance abuse--and early motherhood (Azor & Vectore, 2008).
The damage suffered while living with their families has influence on well-being, however the care provided at the protection system should be compensatory and should help children increase their well-being.
Dell'Aglio and Siqueira (2010) found that stressful events, social support network and family conflicts are predictors of life satisfaction for adolescents at social vulnerability.In addition, the researchers found high scores of stressful events in institutionalized adolescents.Thus, it is possible to think that the level of institutionalized children' well-being may be lower than children living with their families' well-being due to their pre-institutionalization experiences.Other studies with similar population found no significant differences in the well-being of adolescents in care and living with their families.Even though children in care have demonstrated higher means in negative affect, considering, yet, their experiences previous to care (Polleto & Koller, 2011).
The absence of stability and continuity in interpersonal relationships of institutionalized children possibly influence their levels of well-being.In an attempt to diminish these consequences, institutions must take action to preserve family ties (by contacting the nuclear and extended family and keeping the bond with their siblings) and promoting family reintegration (encouraging home visits or at other locations) (Lei n. 12010, 2009).This topic is controversial and widely discussed by several authors (López, Montserrat, Bravo, & Del Valle, 2013) since the contact with the family is not always advisable, neither a predictor of well-being nor of a subsequent family reunification.
In research with institutions of care from southeastern, southern and northeastern Brazil, it was revealed that only 31.2% of institutions encourage family life and 14.1% support family restructuring (IPEA, 2004).In the present study, 35.8% of children reported to have contact with their father, while 66.8% keep in touch with their mother.The highest means of contact with fathers have been reported in periods at home and with mothers on visits.It is notable that children's will is rarely taken into account.This occurs both when they are forced to visit the family who harmed them, as when beneficial contacts with family members are denied.At these matters, the decisions are always from adults.
The items of PWI-SC, GDSI's domains, and single-item OLS significantly discriminated children residing with their families of those who are in residential care, with the exception of the item on "doing things away from home" (PWI-SC) and the neighborhood domain (GDSI).One can understand that the two groups of children (in residential care and who reside with their families) have similar experiences in relation to contact with the neighborhood.According to the Grupo de Trabalho Nacional Pró-Convivência Familiar e Comunitária ( 2009) the physical structure of this modality of care should take into account the architectural pattern of the community in which it is inserted and there should be no nameplate of the house's institutional nature.It is to be located in a residential neighborhood, in order to avoid the stigmatization of its users.
Some reflections can still be made regarding the other items of the scales that discriminate the two groups in favor of children who reside with their families.Satisfaction with the family and home is the item that contributes the most to discriminate groups.Fávero, Vitale and Baptista (2008) argue that one of the causes for instability of the families whose children are living in residential care is the search -especially of mothers in families headed by for the overcoming of social vulnerability.In this attempt, they often change houses and partners, damaging their children's bond with important territorial components that are essential for their development.Dinisman et al. (2012) found that children and adolescents in residential care tend to have lower stability when compared to those who reside with their families, considering the implications of the changes to which they are subjected.These transitions affect their interpersonal relationships, their school, and their leisure activities during free time and even the health services they use, having a negative influence on their well-being.Those children and adolescents whose changes were less frequent because they are living with their families, have higher levels of well-being.In the present study, similar results were found as satisfaction with the school, with time management, to health and safety discriminated groups in favor of children who reside with their families.
For Harden (2004) family stability can be seen as a process of care practices that facilitates and promotes the healthy development of children.The author emphasizes that the experiences previous to residential care were traumatic and that the institution should not be presented as an experience of more instability and insecurity.Furthermore, it emphasizes that the knowledge about the stages of human development is crucial for the educators' understanding of the type of care that a child of a given age needs.
Another item that discriminates in favor of children living with their families regards things that children have.It is understood that the residential care presents access to a more suitable physical structure to their development than families of origin.However even offering access to material goods which they could not have in their families, the instability of the situation and the need to share items in the house can influence their perception of what is of personal use.Children who live with their families, however, probably have more access to items that are of their exclusive use.
Children feel a more elevated sense of belonging to the community, which can be verified trough the nondiscrimination of item and domain concerning community.Even though, regarding the results on the domain personal satisfaction and the item satisfaction with what may happen in the future, studies indicate that passing by residential care situation implies a stigma of exclusion over their lives (Montserrat, et al., 2011;Oliveira & Milnitsky-Sapiro, 2007).The traumatic experiences of the past, the perception of a derogatory social context and the rare opportunities for this population may be part of the explanation for future expectations of these children.In addition, the goal to be fulfilled by institutions is the offer of a quality care while they are unable to return to their families, and to provide educational and social opportunities.However, the perception that the residential care institutions will not be constant and will also not be alternatives they can rely on for the rest of their lives can influence their future prospects (Oliveira & Milnitsky-Sapiro, 2007) and it is this sense of temporarily that affects and prevents the perception of a stable life (Montserrat, 2012).
Another aspect that may be related to the difference between the well-being of children who live with their families and children in residential care may be connected with their future prospects and the "satisfaction with the things they want to be good" item.Montserrat, Casas and Malo (2012) in a study of former residents of foster institutions found that these adults had greater difficulties to continue their formal education than those adults who had lived their childhoods with their families.It is understood that the prospect of being deprived of continuing their studies can have a negative effect on their satisfaction with the future considering that education is an aspect that should be priority for protection systems, since training can open doors to social inclusion.
The study on the lives of children in residential care is very important to evaluate the protective measures to the child whose care is responsibility of the state.Only by a better understanding of their development context it is possible to design and implement public policies that meet the real demands of this population.

Final considerations
This study aimed to identify the well-being of children in residential care and compare it with that of children living with their families.The study of children's satisfaction about service and aspects of their lives, their interests and activities are important for the understanding of wellbeing as a whole.This is still an initial step to meet positive characteristics about these children and to discourage the stereotype of social exclusion to which they are subjected.
The main results of this study indicate that the means of well-being of children residing with their families are significantly higher for the three measures of well-being used when compared with children in residential care.In addition, all items of the measures used discriminated significantly between children residing with their families and those in residential care.It is argued that this result may reflect differences between those children who may be related to the stability of their homes.It is considered that children in residential care are submitted to a great amount of changes over short periods of time what may reflect on their life evaluations and implicate on differences on their well-being.Future studies could investigate the differences in levels of well-being of children who are in more stable foster modalities, such as foster care or foster families.Another alternative that might work well for the well-being of children who are unable to live with their parents, is the placement of children in extended families, with other relatives who can provide them with proper care.This should happen with legally recognition of the public administration and these families should count on psychosocial support (Montserrat, 2012).
In other investigations the stories of children prior to the situation of residential care may be taken into account and enlarge the knowledge of families.In this way it would be possible to identify aspects that contribute to the well-being of this population, enabling the understanding of the influence of residential care at their well-being.
As a limitation of the study, it is emphasized that the measures used may be more sensitive to aspects of children living with their families, because they probably have not been subject to the adversities that children in residential care have been.In this way, the development of measures that take into account the specific conditions of life of those children could be more sensitive to aspects that might influence their well-being.
-Satisfaction with doing things away from home .010

Table 1 .
specific data concerning children in care are presented.Mean and Standard Deviation of the Period in Residential Care, Sex, Age, Siblings in care, contact the Father, Contact with Mother

Table 2 .
Means for items and domains of well-being measures per group

Table 3 .
ANOVA for differences in measures of well-being among children in residential care and children living with their families The discriminant function obtained is significant (χ2 (15) = 61.73,p<.001)andpresents a canonical discriminant correlation of .564.The centroid for the group of children who live with their families is .682andforchildren in residential care is -.674.A Wilk's Lamba of .682wasobtained,which is equivalent to an explained variance by the discriminant function of 32.8%.With regard to classification, the obtained function classified correctly 71.9% of participants in respective groups, considering it an acceptable level of predictive across the profile of the classification function obtained (Table4).