Relationship between Social Support, Life Satisfaction and Subjective Well-Being in Brazilian Adolescents*
Relación entre soporte social, la satisfacción de vida y bienestar subjetivo en adolescentes brasileros
Jorge Castellá Sarriera**
Daniel Abs Tiago Calza
Universidade Federal do Rio Grande do Sul, Porto Alegre, Brasil
Universidade de Girona, UdG, Girona, España
*CAPES - Coordenação de Aperfeicoamento de Pessoal de Nivel Superior y CNPQ - Conselho Nacional de Desenvolvimento Científico e Tecnológico.
**Professor del Instituto de Psicologia y del programa de Post-grado en Psicologia. Coordenador del grupo de Investigación en Psicologia Comunitaria. E-mail: email@example.com
***Universidade Federal do Rio Grande do Sul, Porto Alegre, Brasil. E-mails: firstname.lastname@example.org , email@example.com, firstname.lastname@example.org
Recibido: julio 11 de 2012 | Revisado: agosto 6 de 2014 | Aceptado: marzo 9 de 2015
Para citar este artículo
Sarriera, J. C, Bedin, L., Abs, D., Calza, T., & Casas, F. (2015). Relationship between social support, life satisfaction and subjective well-being in Brazilian adolescents. Universitas Psychologica, 14(2), 459-474. http://dx.doi.org.10.11144/Javeriana.upsy14-2.rbss
Este estudio explora el bienestar subjetivo, la satisfacción general de vida y el apoyo social percibido de los adolescentes brasileños mediante el análisis de diferentes modelos que explican la relación entre estas tres construcciónes. La muestra se compone de 1.588 adolescentes (12 a 16 años, 65,5% mujeres), que contestaron el Indice personal de Bienestar (PWI), una medida de apoyo social percibido de la familia y amigos, y también un solo elemento de satisfacción general Vida escala (OLS). Se realizaron análisis de MANOVA y ANOVA para evaluar las diferencias entre los grupos, con respecto al género, se observa que las niñas presentan medios más bajos que los chicos de la MCO. Se presentó una disminución significativa en la media de la PWI con la edad. Los niños presentan medios más bajos de apoyo social de amigos. El modelo de Ecuaciones Estructurales se utilizó para probar los marcos teóricos sobre los datos. Después de analizar los diferentes modelos, el mejor ajuste del modelo se presenta con el PWI como un factor mediador entre el apoyo social y la satisfacción con la vida. La percepción de apoyo social de la familia y amigos resulta ser una dimensión importante de bienestar entre los adolescentes brasileños. La importancia del apoyo social para el bienestar subjetivo se explica por la identificación de la forma en que contribuye a la satisfacción con la vida en general de los adolescentes.
Palabras clave: adolescencia; bienestar; satisfacción de vida; soporte social
This study explores the subjective well-being, overall life satisfaction and perceived social support of Brazilian adolescents by analysing different models that explain the relationships among these three constructs. The sample is comprised of 1,588 adolescents (aged 12-16, 65.5% female), who answered the Personal Well-Being Index (PWI), a measure of Perceived Social Support of Family and Friends, and also an Overall Life Satisfaction single-item scale (OLS). MANOVA and ANOVA were used to assess the differences between groups, with respect to gender, it is observed that girls present lower means than boys for the OLS. Considering age, there was a significant decrease in the mean of the PWI with age. Boys present lower means of friendssocial support. Structural Equation Modelling was used to test theoretical frameworks on data. After analysing different models, the best fit model is presented using the PWI as a mediating factor between social support and satisfaction with life. The perception of social support from family and friends results to be a significant dimension of well-being among Brazilian adolescents. The importance of social support for the subjective well-being is explained by identifying how it contributes to the overall life satisfaction of adolescents.
Keywords: adolescence; well-being; life satisfaction; social support
The relationship between different well-being measures with a single item of overall life satisfaction (OLS) is very well established in the scientific literature (Cummins, 1998; Cummins et al., 2004), as it is the relationship between different social support measures with both the OLS and different well-being scales in the adult and adolescent populations (Diener, 1984; Huebner, 2004; Sarriera, Abs, Casas, & Bedin, 2012).
However, the relationship between social support and well-being indicators has very different connections depending on the instruments used, the social context where the questionnaires have been administered, cultural differences and other facts. The variations are even more outstanding when research is focused on adolescents (Fernández del Valle & Bravo, 2000).
Some authors have proposed that these three, together with other positive constructs, may form part of a supra-construct, i.e., belong to a same higher order construct (Diener, Suh, Lucas, & Smith, 1999; Stones & Kozma, 1985). However, other authors have proposed that Social Support is a mediator of the adolescentssubjective well-being (Fernández del Valle & Bravo, 2000).
Social support has been defined as a set of inputs of different types, ranging from emotional to material aspects that a person receives from the social network in which he or she is inserted (Gracia, Herrero, & Musitu, 1995). That definition includes structural and functional aspects of social support, related to the perceptions of each subject. The study by Collins, Coffey and Morris (2010) with social work students regarding stress, support and well-being, employs social support measures that assess the relationship of students with its social context. The authors did not find significant differences in gender, age or the year of study. Gayman, Turner, Cislo and Eliassen (2010) assessed the family experiences of 1,267 American adolescents and found them related to low perceived social support from friends and family, especially the negative experiences.
In this study, we assume the concept as proposed by Cobb (1976), later used by Vaux (1988) in the development of the scale of Social Support Appraisals (SSA), defined as the information that makes someone believe he's cared and loved, esteemed and valued and that he or she is a part of a network of mutual obligations. Regarding the first aspect, information of support can be transmitted through intimate situations involving mutual trust and it is often referred to as emotional support. In the second aspect, information of esteem and personal value is transmitted effectively in the presence of others and makes the individual reaffirm his or her sense of personal worth; it can be referred to as recognition.
The perception of social support in adolescence can vary with age, since changes in relationship groups, such as with parents and friends, often take place. This influence has fundamental importance in this stage of life, due to the processes of independence and autonomy, especially in regards to emotional support, loyalty, understanding and intimacy (Antunes & Fontaine, 2005; Lekes, Gingras, Philippe, Koestner, & Fang, 2010). The trust of family and the trust of friends, both included in this analysis, are important to aid the development of the adolescent's independence and autonomy.
This study aims at exploring the following questions: 1) The characteristics of subjective well-being and perceived social support in a Brazilian population of adolescents; 2) The psychometric properties of the three studied measures; 3) The relationship between these three constructs by testing different Structural Equations Models (SEM).
The sample is comprised by 1,588 participants from five Brazilian cities in different regions of the state of Rio Grande do Sul: Porto Alegre, Santa Cruz do Sul, Santa Maria, Rio Grande and Passo Fundo. The characteristics of the sample in terms of age and gender are shown in Table 1. The Students are aged 12 to 16 years old, with an average of 14.13 (SD = 1.26). The sample includes many more girls than boys due to the ethical requirements on the sampling procedure in Brazil, where explicit consent terms from parents are needed. Since girls usually are more proactive in bringing the consent back to school, girlsparents signed more informed consents.
The participants were contacted through public and private schools in the sample cities. The selected schools were randomized from a list provided by Rio Grande do Sul's Department of Education, having as inclusion criteria: 1) to belong to the municipalities in the study, 2) to provide both primary and secondary education and 3) to agree with the participation by signing an institutional agreement. The final participants of the survey were volunteer students who returned a term of free and clear consent signed by them and by their parents or guardians. The questionnaires were applied collectively in a room provided by each school. Two researchers were present during the administration and clarified any questions that arose. All ethical requirements for research with human beings were followed.
Personal Well-being Index (PWI) from Cummins, Eckersley, Pallant, Van Vugt and Misajon (2003) was employed to assess subjective well-being. The scale has 7 agree-disagree type items ranging from 0 to 10 that are related to satisfaction. The PWI is a one-dimensional scale consists of seven domains: health, with living standards, with what one has achieved in life, with security, with the groups of people one is part of, with security about the future and with relations with others. Cummins (1998) found alphas between 0.7 and 0.8 in 16 surveys conducted with the Australian population. Also, the study by Cummins et al. (2004) showed good internal consistency. In the current study with Brazilian adolescents (n = 1.588), the scale showed good internal consistency with alphα = 0.78. The PWI was also tested on 12 to 16-year-old adolescents by Casas, Sarriera, Abs et al. (2012), Casas, Sarriera, Alfaro et al. (2012), and Sarriera et al. (2012).
Overall Life Satisfaction (OLS) single-item scale was used (Casas, 2010 as formulated by Campbell, Converse, & Rodgers, 1976) to measure life satisfaction, consisting in the question "At present, to which extent are you satisfied with your life as a whole?" and uses a range from 0 (completely unsatisfied) to 10 (completely unsatisfied).
Social Support Appraisals Scale (SSA) from Vaux et al. (1986) was employed to assess social support; it measures the adolescentsperception of the social support provided by their family, friends and others in general. In its original form, there are 23 five-point Likert-type items that belong to three dimensions: family, friends and others. In this study, only items related to the family and friendssubscales are used, totalling 12 agree-disagree items with scores ranging from 0 to 10, in order to improve its sensitivity. The 12 items are described in Table 8, and there are six items for each dimension.
Vaux et al. (1986) obtained results with five samples of college students and five samples in communities that showed a high SSA's reliability and validity. The scales had good internal sample consistency; for the sample of college students, the Cronbach's alpha was 0.9 for friends and 0.8 for family. The alpha was 0.9 for friends and 0.81 for family in the community sample. The scales also showed stability after a six-week interval in a sample of college students (r = 0.8 and r = 0.71 for the subscales family and friends). In this paper with Brazilian sample (n = 1.588), the scale also showed good internal consistency, with a Cronbach's alphα = 0.86.
In order to verify the characteristics of subjective well-being and perceived social support in the sample, data were submitted to univariate analysis to assess differences in the subgroups of age and gender. In addition, two multivariate analysis of variance (MANOVA) and univariate analysis of variance (ANOVA) were performed to test for mean differences. In the first, the PWI and the OLS were considered as dependent variables and age and gender as independent variables, and, in the second, the two factors of the SSA scale (family and friends) were the dependent variables, whereas the independent variables remained unchanged.
At a second step, a structural equation modelling (SEM) was estimated with the following steps: 1) Normality and absence of homoscedasticity were tested; 2) Pearson's correlations between variables
were calculated; 3) Factorial Confirmatory Analysis was conducted with the SSA and PWI scales; 4) Models were constructed and adjusted based on theoretical literature; 5) Models were tested for mul-tigroup invariance using the age and gender subgroups (Batista-Foguet & Coenders, 2000; Byrne, 2010; Hair, Anderson, Tatham, & Black, 2005).
The interpretations of results for the SEM consider the Goodness of Fit criterion (Batista-Foguet & Coenders, 2000). Kline (1991) also points out other indicators that verify the adequacy of the model: in order of priority, the Chi-square and the Bentler-Bonnet Normed and Nonnormed Fit Index (NFI, NNFI). This study follows Batista-Foguet and Coenders (2000), who recommend the comparative fit index of Bentler (CFI) and the root mean square error of approximation (RMSEA). Model fit was evaluated according the following criteria: NFI, CFI, and TLI values above 0.95 and RMSEA values below 0.08 (see, Batista-Foguet & Coenders 2000; Byrne 2010).
For the first step, univariate analysis were performed to verify the characteristics of subjective well-being and perceived social support in the sample. Subgroups were constructed according to the adolescentsage and gender. The means and standard deviations of well-being and life satisfaction are shown in Table 2, by subgroups. The means of the PWI were calculated through the arithmetic average of its items on 100.
AdolescentsSocial Support Appraisals means are shown in Table 3. They were measured through the arithmetic average of all items in the SSA sub-scales in this analysis. The girlsaverages are higher than the boysand they are decreasing with age until 15 and then increasing when 16. Boysaverages are more heterogeneous within the age group.
Considering the overall sample, well-being and life satisfaction had the highest average among the youngest adolescents, decreasing with age, even though a small average increase in 16 year-olds can be seen.
Tests of Multivariate Analysis of Variance (MANOVA) and Analysis of Variance (ANOVA)
Tests of multivariate analysis of variance (MANOVA) and Analysis of Variance (ANOVA) were applied by age and gender to verify mean differences in personal well-being, life satisfaction and social support (family and friends). Table 4 shows the results of multivariate analysis of variance for the PWI and OLS. There are significant differences within the age and gender subgroups for both dependent variables and its interaction term.
Table 5 shows the results of the analysis of variance for the PWI and OLS for age, gender and its interaction term. The results of the PWI and OLS analysis of variance (ANOVA) indicated a significant overall difference in the interaction term of gender and age for the PWI, but not for the OLS. In addition, there are significant differences regarding the age group, with a decrease of the PWI mean with age, but not in the OLS. With respect to the gender group, the means are significant higher for boys than for girls for the OLS, but not for the PWI (see Table 2 for means).
For Social Support Appraisals two variables were generated: the social support from family (Family SS) and the social support from friends (Friends SS), by computing the arithmetic average of each group. Table 6 shows the results of multi-variate analysis of variance for the Social Support Appraisals from family and friends. There are significant differences in gender and in the interaction term of gender and age, but not in age by itself.
Table 7 below shows the results of the SSA's (family and friends) analysis of variance by age, gender and its interaction term. The results of the SSA from family and friendsANOVA indicated an overall significant difference in the gender and age interaction term only with respect to social support from friends, and girls present higher means than boys in all ages, except at sixteen (see Figure 4). Separately, only gender had significant differences when regarding social support from friends. There are no significant differences by age, gender and the interaction term in the social support from family.
The next figures show the interaction between age and gender subgroups in the studied variables. Figure 1 shows the graph of the interaction resulting from the personal well-being means, as shown in Table 5, this interaction is significant, and it is possible to see that the PWI means decrease with age for girls but not for boys.
Figure 2 shows the interaction with respect to the Overall Life Satisfaction indicator for age and gender. The means also decrease with age for girls, but as shown in Table 5 this interaction was not significant.
Figure 3 shows the interaction for the perceived social support from family for age and gender. As presented before, this interaction does not present significant differences.
Figure 4 shows the interaction for the perceived social support from friends. Girls present higher means than boys in all ages, except at sixteen years old, when the means are the same.
Structural Equation Models
The skewness and kurtosis of the variables ranged from -3 to +3, what is acceptable according to literature (Byrne, 2010). Levene's test showed significant differences in the dependent variables across subgroups. Pearson correlations were calculated between the studied variables. Significant correlations were found between Overall Life Satisfaction and the Personal Well-Being Index with all items of perceived social support from friends and family and its factors (p < 0.001) (Table 8). Similarly, a significant correlation was found between overall life satisfaction and the personal well-being index (r = 0.52, p < 0.001). The perceived social support items that showed higher correlations with OLS and also with the PWI were family's esteem, admiration, trust and respect (respectively r = 0.43, p < 0.001; r = 0.43, p < 0.001; r = 0.43, p < 0.001; r = 0.41, p < 0.001 for the personal well-being index and r = 0.32, p < 0.001; r = 0.32, p < 0.001; r = 0.32, p < 0.001; r = 0.34, p < 0.001 for the overall life satisfaction variable).
At a second step, a Confirmatory Factor Analysis of the SSA and PWI scales with the pooled sample was applied. Fit statistics showed satisfactory results according to the literature (Batista-Foguet & Coenders, 2000; Byrne, 2010). Results are reported in Table 9.
The next stage consisted on the construction and assessment of two different models. The first is a direct relationship between perceived social support and the OLS single item. Adjustments were applied to this model, which is shown in Figure 5. Subsequently, different specifications relating OLS to SSA and PWI were explored. As a result, the model that displayed the best fit statistics consisted on the Personal Well-Being Index as a mediating factor between social support and satisfaction with life, and is shown in Figure 6. In both models the Family Social Support has a stronger relation with the well-being for adolescents.
The model in Figure 6 was tested for multigroup invariance using age and gender. It was taken into account the comparative fit indices between the models, considering the difference between the CFI of one model to another. The difference between the CFI (ACFI) should not be greater than .01, and thus, the results indicate the possibility of restricting the factor loadings to be the same in the different groups (Milfont & Fisher, 2010). The results are satisfactory with regards to the stability of the model in those two subgroups, as reported by Table 9.
The standardized regression weight estimates for age and gender in Model 2 are shown in Table 10. The data indicate that the PWI and OLS's regression weights reach almost 0.7 at the age of 16. The latent variable Family has a bigger weight than Friends with regards to the PWI. For boys and girls the item that is more related to the latent variable PWI is achievements in life, followed by safety. The same occurs at the ages of 12, 15 and 16.
By assessing the results in MANOVA, the existence of a latent factor that combines the variables OLS and PWI initially stands out. This latent factor, possibly referring to well-being, satisfaction and quality of life issues, shows differences within the gender and age subgroups. However, when the two variables are taking into account separately, other results emerge. The first refers to the gender dissimilarities found when the OLS item is considered alone. The main component of this item is an overall assessment of life satisfaction, which in some cases may suggest a more abstract assessment of life as a whole. In this sense, girls responded differently from boys, with lower scores in average. The difference found suggest that the boysassessment of life is perhaps more optimistic.
The second result regards the PWI variable. Unlike OLS, it consists on a more contextual assessment with detailed scopes that compose an evaluation of personal well-being. Considering the PWI, significant differences were found with age, suggesting that the assessment on the scope and context of life changes during adolescence. In addition to that, there is evidence that it's a more sensitive measure to the age of the adolescent, but not to the gender.
These two results demonstrate the need of a deeper discussion about the quality, specificity and sensitivity of the instruments used to measure well-being or its dimensions. It was found that the OLS variable, perhaps because it is a streamlined item and reports to a global evaluation, is more stable and less susceptible to the adolescent's age, although it lacks specificity. The PWI presented itself as an instrument less susceptible to gender differences and cultural and biological characteristics, since it requires a less abstract evaluation about well-being, encouraging respondents to assign different evaluations in each item.
Regarding the SE Models, the observed relationship between well-being variables (PWI and OLS) and social support variables initially stands out. A specification that considers OLS directly related to social support was analysed. The beta weights suggest that other factors are mediating social support altogether with OLS. The hypothesis employed is that a more contextual evaluation may be closer to the social support assessment. In this case, the personal well-being items can represent a mediation of well-being dimensions with overall life satisfaction and social support.
One of the contributions of this study is the assessment of the relations between perceived support from adolescentsfriends and family and their well-being in a theoretical model that can be applied to different age groups, regardless of gender. The proper fit achieved in this model is important for future research on adolescentsrelationships, social contexts and their direct and indirect impact on adolescentsquality of life. However, despite the good fit of the model, cultural differences can play a role on the results. More research can be developed in this regard, since the model's relations tend to be established differently in other cultures.
From the data analysis we can conclude that the perception of social support from family and friends turns out to be an important dimension of well-being among Brazilian adolescents. Family and friends still have a central role in the adolescent's development (Godinho, Schelp, Parada, & Bertoncello, 2000; Lekes et al., 2010; Pratta & Santos, 2007) and interventions that promote and strengthen the social support from these groups are important. The support system may also develop beyond family and friends, and may constitute an important source of esteem, care, respect and affiliation to a group (Gracia, 1998; López-Cabanas & Chacón, 1997).
Social support is a broad construct and the assessment of perceived support is only one dimension of it (Vaux & Wood, 1987). Therefore, future research could assess other dimensions of this construct, such as the resource network and the supportive behaviour of a given population. Moreover, other factors that predict well-being in adolescents could be analyzed in order to contribute to a broader understanding of how this important construct is configured. This study has limitations, one is the fact that we have only cross-sectional data to check for differences by age. Longitudinal data could provide better evidence of change of well-being over time for adolescents. Furthermore, the difference in the number of participants by gender and age may have influenced the results, so more research is needed with larger samples for age and also with more male participants.
Comparative studies that determine the parentsperception of the adolescentssupport and well-being is recommended, since support from family is critical in predicting the adolescent well-being in most of the factors analyzed. Furthermore, guidelines for family interventions can be helpful in order to reinforce the adolescent's social support perception and to improve their level of well-being. Additionally, considering the findings of this paper, further research on the differences between perceived support and the adolescentswell-being taking into account gender and age is also suggested, since it also plays a role on adolescentsperceptions.
Antunes, C., & Fountaine, A. M. (2005). Percepgáo de apoio social na adolescencia: análise fatorial confirmatória da Escala Social Support Appraisals. Paidéia, 15(32), 355-366.
Batista-Foguet, J. M., & Coenders, G. (2000). Modelos de ecuaciones estructurales. Madrid: La Muralla.
Byrne, B. M. (2010). Structural equation modeling with AMOS: Basic concepts, applications, and programming (2nd ed.). New York: Routledge.
Campbell, A., Converse, P. E., & Rogers, W. L. (1976). The quality of American life: Perceptions, evaluations, and satisfactions. New York: Russell Sage.
Casas, F. (2010). El bienestar personal: su investigación en la infancia y la adolescencia. Encuentros en Psicología Social, 5(1), 85-101.
Casas, F., Sarriera, J. C., Abs, D., Coenders, G., Alfaro, J., Saforcada, E., & Tonon, G. (2012). Subjective indicators of personal well-being among adolescents. Performance and results for different scales in Latin-language speaking countries: A contribution to the international debate. Child Indicators Research, 5(1), 1-28.
Casas, F., Sarriera, J., Alfaro, J., González, M., Malo, S., Bertran, I., ... Valdenegro, B. (2012). Testing the personal wellbeing index on 12-16 year-old adolescents in 3 different countries with 2 new items. Social Indicators Research, 105(3), 461-482.
Cobb, S. (1976). Social support as a moderator of life stress. Psychosomatic Medicine, 38(5), 300-314.
Collins, S., Coffey, M., & Morris, L. (2010). Social work students: Stress, support and well-being. Bristish Journal of Social Work, 40(1), 963-982.
Cummins, R. A. (1998). The second approximation to an international standard for life satisfaction. Social Indicators Research, 43(3), 307-334.
Cummins, R. A., Eckersley, R., Pallant, J., Van Vugt, J., & Misajon, R. (2003). Developing a national index of subjective wellbeing: The Australian Unity Wellbeing Index. Social Indicators Research, 64, 159-190.
Cummins, R. A., Eckersley, R., Lo, S. K., Okerstrom, E., Hunter, B., & Davern, M. (2004). Australian Unity Wellbeing Index: Cumulative psychometric record. Melbourne: Australian Centre on Quality of Life, School of Psychology, Deakin University. http:// acqol.deakin.edu.au/index_wellbeing/index.htm
Diener, E. (1984). Subjective well-being. Psychological Bulletin, 95(3), 542-575.
Diener, E., Suh, E. M., Lucas, R. E., & Smith, H. L. (1999). Subjective well-being: Three decades of progress. Psychological Bulletin, 125(2), 276-302.
Fernández del Valle, J., & Bravo, A. (2000). Estructura y dimensiones de apoyo de la red social en los adolescentes. Anuario de Psicología, 31 (2), 87-105.
Gayman, M. D., Turner, R. J., Cislo, A. M., & Eliassen, A. H. (2010). Early adolescent family experiences and perceived social support in young adulthood. Journal of Early Adolescence, 31 (6), 880-908. doi: 10.1177/0272431610376247
Godinho, R. A., Schelp, J. R. B., Parada, C. M. G. L., & Bertoncelo, N. M. F. (2000). Adolescentes e grávidas: onde buscam apoio? Revista Latino-Americana de Enfermagem, 8(2), 25-32. doi: 10.1590/S010411692000000200005
Gracia, E. (1998). El apoyo social en la intervención comunitaria. Barcelona: Paidós.
Gracia, E., Herrero. J., & Musitu, G. (1995). El apoyo social. Barcelona: PPU.
Hair, J. F., Anderson, R. E., Tatham, R. L., & Black, W. C. (2005). Análise Multivariada de dados. Porto Alegre: Bookman.
Huebner, E. S. (2004). Research on assessment of life satisfaction of children and adolescents. Social Indicators Research, 66, 3-33.
Kline, R. B. (1991). Latent variable path analysis in clinical research: A beginner's tour guide. Journal of Clinical Psychology, 47(4), 471-484.
Lekes, N., Gingras, I., Philippe, F., Koestner, R., & Fang, J. (2010). Parental autonomy-support, intrinsic life goals, and well-being among adolescents in China and North America. Journal of Youth and Adolescence, 39(8), 858-869. doi: 10.1007/s10964009-9451-7
López-Cabanas, M., & Chacón, F. (1997). Apoyo social, redes sociales e grupos de autoayuda. In M. López-Cabanas & F. Chacón (Eds.), Intervención psicosocial y servicios sociales: un enfoque participa-tivo (pp. 183-215). Madrid: Síntesis Psicológica.
Milfont, T. L., & Fischer, R. (2010). Testing measurement invariance across groups: Applications in cross-cultural research. International Journal of Psychological Research, 3(1), 111-121.
Pratta, E. M. M., & Santos, M. A. (2007). Família e adolescencia: A influencia do contexto familiar no desenvolvimento psicológico de seus membros. Psicologia em Estudo, 12(2), 247-256.
Sarriera, J. C., Abs, D., Casas, F., & Bedin, L. (2012). Relations between media, perceived social support and personal well-being in adolescence. Social Indicators Research, 196 (3), 545-561.
Stones, M. J., & Kozma, A. (1985). Structural relationships among happiness scales: A second order factorial study. Social Indicators Research, 17, 19-28.
Vaux, A. (1988). Social support: Theory, research, and intervention. New York: Praeger.
Vaux, A., & Wood, J. (1987). Social support resources, behaviour, and appraisals: A path analysis. Social Behavior and Personality, 15(1), 105-109.
Vaux, A., Phillips, J., Thomson, B., Holly, L., Williams, D., & Stewart, D. (1986). The Social Support Perceptions (SSA) Scale: Studies of reliability and validity. American Journal of Community Psychology, 14, 195-220.