Resilience in Socioeconomically Disadvantaged Turkish Adolescents: An Ecological Perspective

Abstract

The purpose of this study was to examine the role of self-esteem as mediator in the relationships between perceived parental acceptance/involvement, perceived peer social support, sense of school belonging and resilience in adolescents attending schools located in low socioeconomic districts. The sample of the study consisted of 1312 high school students (673 female, 639 male) between the ages of 13 to 19 (M = 15.67, SD = 1.18). Structural equation modeling was conducted to test the hypothesized model. Results showed that perceived parental acceptance/involvement, perceived peer social support, sense of school belonging and self-esteem were positive and significant predictors of resilience. Furthermore, self-esteem partially mediated the association between perceived parental acceptance/involvement, perceived peer social support, sense of school belonging and resilience. The proposed model explained 33% of the variance in resilience. Overall, findings contributed to the understanding of the simultaneous influence of multilevel resources in adolescent resilience.

Keywords

resilience adolescence parental involvement/involvement school belonging peer social support self-esteem ecological perspective

Introduction

Adolescence is a developmental period in which various changes and challenges bring stress and pressure to young people (Steinberg et al., 2011). While the majority of adolescents pass through without enduring problems (Eccles et al., 1993), adolescents with specific disadvantages may be thought vulnerable due to additional stress or lack of resources. Resilience research considering how individuals successfully adapt to adversities could aid understanding of protective factors relevant to these adolescents (Prince-Embury & Saklofske, 2014).

What resilience and related concepts such as protective and risk factors mean? The concept of resilience has been defined in various ways. Some researchers understand resilience as a process, capacity or outcome of positive adaptation in the face of significant adversity (Masten et al., 1990), others have conceived it as a trait, individual attribute or constellation of characteristics (Wagnild, 2010). This trait is not fixed but may change over time (Fletcher & Sarkar, 2013).

Personal, environmental or situational characteristics that contribute to positive outcomes in the context of risk or adversity are considered protective factors. Low socio-economic status (SES), disability, chronic illness or traumatic life experiences are common risk factors which negatively affect human development (Masten & Reed, 2002). Interactions between these protective agents could enhance resilience in youth (Stepleman et al., 2009).

The impact of specific demographic variables in adolescent resilience is debatable. Some studies reported higher resilience in girls (Dias & Cadime, 2017; Onat, 2010) while others suggested higher resilience in boys (Dayıoğlu, 2008) or showed no gender difference (Aydın-Sünbül, 2016; Sagone & De Caroli, 2014). The relationship between age and resilience was positive in some studies (Atik, 2013; van Harmelen et al., 2017), negative (Onat, 2010) or unrelated (Dias & Cadime, 2017; Gündaş, 2013) in others. Similarly, studies of adolescents found both a positive relationship (Uçanok & Güre, 2012) and no association (Atik, 2013; van Harmelen et al., 2017) between resilience and SES.

In recent years, resilience research has underlined the significance and necessity of examining contextual variables and interactions between protective factors. Many investigations of relations between protective agents or resources in resilience have adopted the ecological system approach (Ungar, 2012). Bronfenbrenner’s (1979) bio-social-ecological system model of human development, which guides the theoretical framework of this study, proposes that the simultaneous influence of individual, interpersonal and contextual factors shape a child’s development. This approach also highlights the power of proximal contexts (e.g., individual, family, and peer factors in microsystem level) in developmental outcomes. That’s to say, examination of individual, parental, peer and school related resources coupled with the study of interactions between these micro-systemic domains conducted from an ecological system perspective could help discover complex mechanisms and interactions in youth resilience.

Parental Acceptance/Involvement

Lamborn et al. (1991) defined parental acceptance/involvement as the degree to which an individual perceives his/her parents as caring, responsive and involved. Adolescents who perceive their parents as supportive or caring were found to have enhanced resilience in the face of negative life conditions (Cauce et al., 2013). Child socialization theory asserts that connection, which is conveyed through consistent, warm and positive relations between child and parent, facilitates children in attaining coping skills (Barber, 1997). Similarly, perceived parental acceptance and involvement predicted resilience or adaptive coping with stressful life events in adolescents (Zakeri et al., 2010).

Sense of School Belonging

Sense of school belonging refers to the extent to which students perceive personally valued, accepted or included by others in the school (Goodenow, 1993). Schools are filled with networks which could satisfy the need of adolescents to belong, who have a developmental need for deeper relationships with peers and adults outside their families (Benard, 2004). The belonging hypothesis suggests that individuals who are satisfied in their need to belong through supportive relationships have a sense of assistance and support in coping with adversity (Baumeister & Leary, 1995). Although the influence of school belonging has been under-researched (Prince-Embury & Saklofske, 2014), some studies revealed that sense of school belonging was positively associated with resilient outcomes (Kia-Keating & Ellis, 2007; Nuttman-Shwartz, 2018) or mental health in adolescents with negative life experiences (van Ryzin et al., 2009). Sense of school belonging may be especially important for youth with disadvantageous socioeconomic conditions, because socioeconomic disadvantage brings a decreased sense of school belonging (Chiu et al., 2016).

Peer Social Support

Social support refers to information which leads people to perceive that they are valued, loved or cared for as member of a social network (Cobb, 1976). With the onset of adolescence, individuals largely rely on their peers for social support (Helsen et al., 2000). Supportive peer relations are one of the societal protective factors that contribute to resilience in adolescents (Olsson et al., 2003). The main effect model posits that social support contributes to wellbeing, self-value and positive affect, regardless of individuals are under stress or not. The buffering model asserts that social support moderates the negative effects of stress and leads to favorable outcomes (Cohen & Wills, 1985). Through the main effect and buffering models, resilience studies have supported the idea that perceived peer social support has been a protective factor for adolescents experiencing challenges such as low SES (Seidman & Peterson, 2003).

The Mediating Role of Self-esteem

Self-esteem, which refers to one’s perceptions about overall self-worth or self-acceptance (Rosenberg, 1965), has been one of the most prominent individual level factors in resilience research (Kumpfer, 1999). The transactional theory of stress and coping asserts that self-esteem may help individuals believe in their sense of control, ability or resources in the face of stress, and thus facilitate adaptive responses (Lazarus & Folkman, 1984). Current research findings also support that self-esteem operates as a protective factor in resilience (Zimmerman & Brenner, 2010).

Theoretical conceptualization of self-esteem suggests that it develops and flourishes within a supportive network of social relations, is internalized as an individual attribute, impacts developmental outcomes (Harter, 2006). Research findings supported adolescent self-esteem’s association with acceptance, caring or perceived parental involvement (Zakeri et al., 2010). Self-esteem in adolescents was also found to be related to schools’ provision of a sense of belonging, another important socialization context (Strudwicke, 2000). The positive association between perceived peer social support and self-esteem is well supported (Colarossi & Eccles, 2003).

Current Study

Considering that perceived parental acceptance/involvement, perceived peer social support, and sense of school belonging is associated with self-esteem, and self-esteem is related to resilience in adolescents, self-esteem may be a possible pathway through which parental, peer and school related agents can cultivate resilience. The hypothesis of this study built on a structural model that included these variables. The specific research question addressed by this study was: “To what extent resilience of adolescents attending schools in low SES districts is explained by the hypothesized structural model, comprising self-esteem, parental acceptance/involvement, peer social support and sense of school belonging?” The hypothesis under investigation included the expectations that (a) parental acceptance/involvement, perceived peer social support and sense of school belonging are directly related to self-esteem and resilience; (b) self-esteem is directly related to resilience; (c) parental acceptance/involvement, perceived peer social support and sense of school belonging have indirect relations to resilience through the mediating effect of self-esteem.

Method

Participants

The sample of this study was composed of 1312 9th, 10th, 11th, and 12th graders (673 female, 639 male) who attend schools located in low SES districts in Istanbul. The context in which these students attend school or reside is characterized by high population density, high numbers of illiterate individuals or elementary school graduates and above-average household sizes. The economic development index is below average. In regard to educational opportunities, the number of students per teacher is on average or higher than the average of 13. The number of students per classroom is also on average or higher than the average of 23. Residents in these districts reported high dissatisfaction with household income, availability and structure of primary medical services and a sense of neighborhood safety (Şeker, 2011; TÜİK, 2013).

Purposive sampling was utilized for the sample selection process. The age of students ranged from 13 to 19 with a mean age of 15.67 (SD = 1.18). Participants’ grade levels were as follows: 373 were 9th graders, 313 were 10th graders, 324 were 11th graders and 302 were 12th graders. The study obtained a high response rate of 91%. Residential district information was also obtained because students do not necessarily reside in low SES regions, even if they attend schools in low SES districts. It was found that, according to SES indicators developed by TÜİK (2013), nearly 95% of participants in the sample dwell in low SES districts.

Measures

14-Item Resilience Scale (RS-14)

RS-14 was developed by Wagnild (2010) as an alternative to the 25-item Resilience Scale used to assess the degree of resilience in individuals. Wagnild and Young (1993) conceptualized resilience as an individual attribute, characteristic or ability to cope with challenges successfully. The original scale was developed by assessing characteristics of successfully adapted women who had experienced a recent stressful experience. The results acknowledged five constituents of resilience: equanimity (i.e., balanced and broad perspective about life experiences), perseverance (i.e., persistence despite adversity), self-reliance (i.e., one’s belief in oneself and efficacy), meaningfulness (i.e., belief in that life has a purpose), and existential aloneness (i.e., realization that each one’s life is unique) (Wagnild & Young, 1993).

RS-14 is scored on a seven-point Likert type scale ranging from “strongly disagree” to “strongly agree,” higher scores indicating a higher level of resilience. Items include statements such as “I usually manage one way or another,” “my belief in myself gets me through hard times” and “when I’m in a difficult situation, I can usually find my way out of it.” The findings of principal component analysis supported a one-factor structure of the original scale. The scale was found to have a Cronbach alpha value of .93 (Wagnild, 2010). The results of confirmatory factor analysis (CFA) yielded a one-factor structure of the Turkish version of RS-14 (χ²/df = 4.4, GFI = 0.94, RMSEA = 0.07, CFI = .93, TLI = .91). The Cronbach alpha was reported at 0.81 (Aydın-Sünbül, 2016). In the current study, CFA findings supported a single factor structure of RS-14 (χ2/df = 5.81, GFI = 0.95, SRMR = 0.04, RMSEA = 0.06 [90% CI = 0.06, 0.07], CFI = 0.95, TLI = 0.94). The Cronbach’s alpha was .90 for the overall scale.

Parental Attitude Scale (PAS)

This scale was developed by Lamborn et al. (1991) to assess individuals’ perceived parental attitudes. It is composed of three subscales; acceptance/involvement, strictness/supervision and psychological autonomy. This study used a parental acceptance/involvement subscale with 18 items, scored on four-point Likert type scale ranging from 1 (not alike at all) to 4 (very much like). Higher scores indicated a higher level of perceived parental acceptance/involvement. The Cronbach alpha coefficient was reported to be .72 (Steinberg et al., 1994). In the Turkish adaptation study, the Cronbach alpha coefficient value was .70, and two-week test-retest reliability coefficient was .82 for this subscale (Yılmaz, 2000). The results of CFA also confirmed a single factor structure of subscale under this study (χ²/df = 3.32, GFI = 0.98, SRMR = 0.02, RMSEA = 0.03 [90% CI = 0.02, .04], CFI = 0.98, TLI = 0.97). A Cronbach alpha coefficient of. 69 was found.

Psychological Sense of School Membership Scale

This scale was developed in order to measure students’ perceived sense of belonging or psychological membership in the school context (Goodenow, 1993). Psychological sense of school membership scale (PSSM) consists of two subscales; a sense of school belonging and a sense of rejection. This study used a sense of school belonging subscale with 13 items. This five-point rating scale ranges from 1 (not true) to 5 (absolutely true) for each item, higher scores indicating a greater sense of school belonging. The original scale, had a Cronbach alpha of .80 overall (Goodenow, 1993). In the Turkish version of the scale, the Cronbach alpha for the sense of school belonging subscale was 0.88 (Sarı, 2013). In this study, CFA findings confirmed the single factor structure of the subscale (χ2/df = 8.41, GFI = 0.94, SRMR = 0.05, RMSEA = 0.08 [90% CI = 0.07, 0.08], CFI = 0.92, TLI = 0.90) and the Cronbach’s alpha was .88.

Social Support Appraisals Scale for Children

Social support appraisals scale for children (SSASC) was developed to assess perceived social support of children and adolescents from their family, teachers and peers (Dubow & Ullman, 1989). The scale includes three factors; peer, teacher, and family social support. Peer social support subscale with 19 items was used in this study. The respondents were asked to rate items on the five-point rating scale ranging from 1 (never) to 5 (always). Higher scores indicated higher perceived peer social support. The Cronbach alpha was .88 for the peer social support subscale (Dubow & Ullman, 1989). In the Turkish adaptation study, the Cronbach alpha was .89 (Gökler, 2007). The results of CFA supported a single factor structure of this subscale for the current study (χ²/df = 8.44, GFI = 0.91, SRMR = 0.06, RMSEA = 0.075 [90% CI = 0.07, 0.08], CFI = 0.91, TLI = 0.88). The Cronbach’s alpha was .89.

Rosenberg Self-Esteem Scale (RSES)

RSES was developed to measure individuals’ positive attitudes towards themselves and the extent to which they perceive themselves as worthy (Rosenberg, 1965). The scale is composed of 10 items scored on a four-point rating scale ranging from 1 (totally right) and 4 (totally wrong). Higher scores indicate higher self-esteem. Rosenberg (1965) reported a Cronbach alpha coefficient of .80. Test-retest reliability with a two-week interval was .85. In the Turkish adaptation study, the Cronbach alpha coefficient was .87, and the test-retest reliability coefficient was .75. A correlation of .71 between RSES scores and psychiatric interviews indicated criterion-related validity of the scale (Çuhadaroğlu, 1985). In the current study, CFA results supported a single factor structure of RSES (χ²/df = 4.44, GFI = 0.98, SRMR = 0.04, RMSEA = 0.05 [90% CI = 0.04, 0.06], CFI = 0.96, TLI = 0.96), and the Cronbach’s alpha was .86.

Data Analysis

Analysis began with examining descriptive statistics about the sample and bivariate correlations between study variables using SPSS 23 software (Table 1). Then, structural equation modeling (SEM) was conducted with AMOS 22 software to test the hypothesized model and evaluate direct and indirect associations between variables. A missing value analysis showed that missing data points were below 5% of the total cells for each variable. The missing data imputation method used an Expectation-Maximization algorithm, which performs estimations for missing values and gives maximum likelihood estimations for parameters in a two-stage process (Tabachnick & Fidell, 2013).

Table 1.

Descriptive Statistics and Intercorrelations among Study Variables.

Variable	M	SD	1	2	3	4
1. Parental acceptance/involvement	24.12	4.26
2. Sense of school belonging	3.47	0.74	0.37**	—
3. Peer social support	65.86	10.46	0.32**	0.49**	—
4. Self-esteem	29.72	5.64	0.35**	0.41**	0.44**	—
5. Resilience	74.04	16.53	0.30**	0.37**	0.37**	0.46**

**p < .01.

Assumptions of SEM were checked prior to inferential analysis. Mardia’s (1975) test indicated that an assumption of multivariate normality was not ensured (p < .001). An item parceling technique was used to mitigate this (Brown, 2015; Kline, 2016). No other assumptions were found to be violated. To examine direct and indirect relationships among variables, the Bootstrapping method of resampling, was used as remedy for multivariate abnormal data, with 2000 bootstrapped samples (Brown, 2015).

The following criteria values were utilized to check goodness of fit values. A non-significant chi-square value (Schumacker & Lomax, 1996), and a normed chi-square value below 3 (Kline, 2016) indicate good model fit. A GFI value of 1.00 indicates a perfect fit, and values higher than 0.90 show good fits (Jöreskog & Sörbom, 1993). A SRMR value of 0.00 indicates a perfect fit while values below 0.08 refer to a good fit (Hu & Bentler, 1999). RMSEA close fit values are lower than 0.05; mediocre fit values between 0.05 and 0.08; and poor fit values higher than 0.10 (Browne & Cudeck, 1993). CFI and TLI values above 0.90 mean a good model fit (Bentler, 1990; Schumacker & Lomax, 1996) while values closer to 1.00 indicate a better model fit (Brown, 2015).

Procedure

Low SES districts were determined by criteria such as income level, education level of parents, population density, unemployment rate, the number of students in schools, migration rate, access to health services and quality of life (Stepleman et al., 2009; TÜİK, 2013). Ethical permissions were obtained. Seven high schools from Sultangazi and Umraniye districts were contacted. After organizing available class hours for data collection, school counselors or teachers were asked to give parent consent forms to students. The administration of scales took approximately fifty minutes. The study also obtained information about gender and residential district of students in this time.

Results

Descriptive Statistics

As seen in Table 1, participants’ average scores for all of the study variables were slightly above the mean. Study variables were positively and significantly associated with each other according to Pearson product-moment correlations coefficients. All associations between study variables were moderate.

Item Parceling

Three parcels for each latent variable were created because three indicators for a construct allow a model to be just-identified, thus decrease estimation bias. A factorial algorithm technique was utilized to build parcels, as it allows the researcher to create balanced parcels (Matsugana, 2008). Factor loadings of items were used to assign them into the parcels. Then, the average of the items in the parcels were used instead of the individual items. The descriptive statistics about parcels and Cronbach alpha values are presented in Table 2.

Table 2.

Descriptive Statistics and Cronbach Alpha of Parcels.

Parcels	M	SD	Skewness	Kurtosis	Cronbach α
Parent1	3.03	0.60	−0.67	0.39	.68
Parent2	2.89	0.68	−0.34	−0.47
Parent3	3.19	0.78	−1.02	0.60
School1	3.29	0.86	−0.39	−0.18	.86
School2	3.60	0.86	−0.53	0.09
School3	3.56	0.80	−0.57	0.19
Peer1	3.97	0.70	−0.77	0.19	.88
Peer2	3.79	0.69	−0.56	−0.03
Peer3	3.88	0.68	−0.64	0.06
Se1	2.97	0.67	−0.37	−0.30	.88
Se2	3.03	0.65	−0.28	−0.60
Se3	2.93	0.58	−0.21	−0.20
Res1	5.33	1.26	−1.10	1.10	.91
Res2	5.15	1.25	−0.86	0.40
Res3	5.41	1.34	−1.16	1.06

Note. Parent = Parental acceptance/involvement; School = Sense of school belonging; Peer = Peer social support; Se = Self-esteem; Res = Resilience.

Measurement Model Testing

Relations among latent variables and indicators were examined through measurement model analysis (Kline, 2016). According to CFA results, the Chi-square statistic was statistically significant (χ² (80) = 278.82, p = .000). However, Brown (2015) suggested that χ² value is easily inflated due to large sample size, therefore significant results are expected. The normed Chi-square value (χ²/df) of 3.49 was slightly above cutoff value of 3. GFI = 0.97, SRMR = 0.03, RMSEA = 0.044 (90% CI = 0.038, 0.049), CFI = 0.98 and TLI = 0.98 values indicated a good fit for the measurement model.

Structural Model Testing

SEM analysis results indicated a statistically significant Chi-square value (χ² (107) = 363.41, p = .000), as expected due to the large sample size (Brown, 2015). The normed Chi-square value (χ²/df) of 3.40 was slightly above the cutoff value of 3 (Kline, 2016). GFI = 0.97, SRMR = 0.03, RMSEA = .043 (90% CI = 0.038, 0.048), CFI = 0.98 and TLI = 0.97 values indicated that the structural model fits the data. Age and residential district variables (dwelling in low socioeconomic district or not) were specified as covariates. No significant relationship was found between resilience and either age (p = .112) or residential district (p = .470).

In order to check whether the structural model would vary by gender, a multiple group SEM analysis was carried out. The results showed a good fit for the model (χ² (214) = 503.021, Bollen-Stine corrected p = .000, χ²/df = 2.35, GFI = .96, SRMR = 0.05, RMSEA = 0.03, CFI = 0.98, TLI = 0.97). In other words, factor covariance and factor invariances among latent variables in the hypothesized model is equivalent across gender groups. In addition, the model comparison showed that the model did not significantly differ between girls and boys, Δχ² (Δdf = 21) = 32.82, p = .05). The squared multiple correlation coefficient values indicated that 33% of the variance in resilience was explained by the overall model. In addition, 35% of the variance in self-esteem was accounted for by parental acceptance/involvement, sense of school belonging and peer social support.

Direct and Indirect Relationships

According to bootstrapping results, all proposed paths were statistically significant, and standardized path coefficients (β) values ranged between .13 and .33 (Figure 1). β values less than .10, around .30, and over .50 referred to small, medium, and large effect size, respectively (Kline, 2016). According to this index, parental acceptance/involvement (β = .13, p < .01), sense of school belonging (β = .14, p < .01), and peer social support (β = .13, p < .01) had a direct and small effect on resilience. The direct effect of parental acceptance/involvement (β = .26, p < .01), sense of school belonging (β = .17, p < .01), and peer social support (β = .30, p < .01) on self-esteem were small to medium. Self-esteem had a direct and medium effect on resilience (β = .33, p < .01). All indirect paths from exogenous variables on the endogenous variable through mediator variables were statistically significant. The indirect effect of parental acceptance/involvement (β = .08, p < .01), sense of school belonging (β = .06, p < .01) and peer social support (β = .10, p < .01) on resilience via self-esteem was small but positive (Table 3).

Figure. 1.

The hypothesized structural model with standardized estimates and significant paths.

Table 3.

Bootstrapped Results of Standardized Total, Direct, and Indirect Effects.

Paths	β	p	BC Interval
Direct Effects
Parental acceptance/involvement → Resilience	.129	.001	(0.059, 0.204)
Parental acceptance/involvement → Self-esteem	.255	.001	(0.173, 0.338)
Sense of school belonging → Resilience	.137	.001	(0.060, 0.225)
Sense of school belonging → Self-esteem	.174	.001	(0.091, 0.248)
Peer social support → Resilience	.125	.001	(0.052, 0.204)
Peer social support → Self-esteem	.298	.001	(0.226, 0.369)
Self-esteem → Resilience	.329	.001	(0.245, 0.401)
Indirect Effects
Parental acceptance/involvement→ Self-esteem→ Resilience	.084	.001	(0.054, 0.119)
Sense of school belonging → Self-esteem → Resilience	.057	.001	(0.029, 0.089)
Peer social support → Self-esteem → Resilience	.098	.001	(0.067, 0.134)
Total Effects
Parental acceptance/involvement → Resilience	.213	.001	(0.140, 0.292)
Sense of school belonging → Resilience	.194	.001	(0.114, 0.280)
Peer social support → Resilience	.224	.001	(0.149, 0.294)

Note. Reported BC intervals are the bias corrected 95% confidence interval of estimates resulting from bootstrap analysis.

Discussion

The aim of this study was to examine predictors of resilience in adolescents attending school in low-SES districts. Based on the ecological system theory, it developed and tested a mediational model in which parental acceptance/involvement, sense of school belonging and peer social support was proposed to predict resilience through self-esteem. SEM analysis results indicated a good fit between the proposed model and the current data and supported the hypothesized relationships between variables.

The proposed model did not differ according to gender. Studies examining resilience across gender are somewhat unclear. It should be kept in mind that resilience is assessed in various ways, thus gender differences in resilience could refer to gender differences in self-efficacy, depression, wellbeing etc., depending on the conceptualization of resilience in any given study. Our findings are similar to findings of some studies using RS-14 scale and reporting no gender differences (Aydın-Sünbül, 2016; Sagone & De Caroli, 2014).

Parental acceptance/involvement had a positive and significant direct effect on resilience. Theories of family socialization suggest that parents adopting an accepting, caring or involved attitude contribute to competence and coping skills (Baumrind, 1991), and thus support resilience in youth (Zakeri et al., 2010). Since socioeconomic problems bring about negative parental attitudes, lowered parental competence or adjustment problems in children (Conger et al., 1997), parental acceptance/involvement may be especially critical for resilience of adolescents attending schools in low SES regions. Parental acceptance/involvement had a positive and significant relationship with self-esteem. This finding is in line with theories of self-esteem which assert that children internalize opinions held by significant figures in their life and that such an internalization mostly determines self-evaluations of the child’s own worth or esteem (Harter, 2006). Parents who embrace an accepting, loving and responsive approach toward their children play an important role in fostering self-esteem (Aydın et al., 2014; Zakeri et al., 2010).

The results showed that sense of school belonging had a positive and significant direct effect on resilience. Schools provide a generous context in which to satisfy an increased need for belonging in the adolescence period, and thus foster positive adaptation and resilience (Ungar, 2012). Considering that socioeconomically disadvantageous adolescents are under a higher risk of decreased sense of school belonging (Chiu et al., 2016; Goodenow, 1993), a social environment in school where students perceive that they are accepted, cared or included may operate as a valuable resource for adolescents receiving education in stressful socioeconomic areas. Sense of school belonging also showed positive and significant association with self-esteem. As the belonging hypothesis suggested (Baumeister & Leary, 1995), belonging or relatedness is one of humankind’s basic psychological needs for psychological health. One of the determinants of psychological health is an individual’s relationship with the self, or self-esteem. Therefore, a school context that satisfies the students’ need to belong may be a valuable resource for the development and maintenance or self-esteem in adolescents (Strudwicke, 2000).

Perceived peer social support had a positive and significant direct effect on resilience. Peers are an important source of social support for adolescents (Helsen et al., 2000). Studies indicate that peer social support has a protective effect against negative life conditions such as low SES (Seidman & Peterson, 2003). In this sense, the perception of being socially supported or accepted by peers may promote resilience in youth experiencing the stress of adolescence years and additional difficulties related to socioeconomic conditions. This study also found a positive significant relationship between perceived peer social support and self-esteem in adolescents. This finding is consistent with self-esteem theories suggesting that children need others’ perspectives or support for consolidation of self-esteem as they grow up (Harter, 2006). Young people who perceive that they are valued, cared for or loved by their peers may have elevated self-esteem while handling challenges or stressors.

A positive significant association between self-esteem and resilience was found. It is well-known that self-esteem functions as an important protective factor in helping individuals get through traumatic or challenging life experiences (Kumpfer, 1999). During adolescence, fluctuation in self-esteem is common (Harter, 2006). This decrease in self-esteem may be more distinct for adolescents who are exposed to low SES disadvantages, because these disadvantages impose risk on the healthy development of self-esteem (Veselska et al., 2010). Thus, self-esteem may be an essential internal resource in the resilience of adolescents exposed to impact of low SES community.

Regarding the indirect effects between study variables, self-esteem was found to be a significant mediator between parental acceptance/involvement and resilience. This finding supports the premise that self-esteem may operate as an individual resource transmitting the positive effects of positive parental attitudes onto resilience. The mediating effect of self-esteem on the relationship between sense of school belonging and resilience was also positive and significant. This finding means that self-esteem may be a significant mechanism in the association between the sense of belonging in a school context and the resilience of adolescents. Self-esteem was also found to be a significant mediator between perceived peer social support and resilience. This finding points out that peer social support may facilitate resilience through the enhancement of self-esteem in adolescents. However, there has been limited research on the relation between self-esteem, peer social support and resilience.

Several implications could be drawn from the findings of this study. This study has potential to provide meaningful information in understanding adolescent resilience from an ecological perspective. The variables in the proposed model explained a particular percentage of variance in resilience. Undoubtedly, other factors predict resilience in adolescents attending schools in low SES regions. This model included only unidirectional pathways between variables. However, models including possible bidirectional pathways among variables could explain two-way influences between resources and resilience. Furthermore, this cross-sectional study revealed associations, not causal effects, among variables at a single point in time. Changes in these associations over time and causal relations between variables could be studied in further research. The recent emphasis on resilience in the ecological perspective promotes examining resilience in a holistic way (Ungar, 2012). Further studies could investigate other protective factors as well as interactions between different systems in adolescent ecology in order to understand complex interactions and mechanisms in resilience. Additionally, similar resilience model studies could be conducted with individuals experiencing different risk factors.

The findings could guide youth resilience prevention and intervention studies. The simultaneous interventions to different microsystemic factors such as peer, school and family environment could be targeted in order to enhance resilience of adolescents. Specifically, social and environmental level interventions addressing the relationship between parents and adolescents, supportive peer relationships and sense of school belonging could be conducted. The individual level interventions such as self-esteem building programs could be designed especially for adolescents with low SES background.

Limitations

The findings should be evaluated in the light of the study’s limitations. Firstly, convenience sampling ensures that the sample of the study was composed of adolescents from low socioeconomic districts, referring to a homogenous group. The lack of randomization in sampling limits the generalizability of results to similar groups. Further studies with adolescents from diverse regions may improve the generalizability of results. Secondly, inaccurate responses are expected in self-report measures. What is more, resilience was assessed as an individual characteristic via a self-report measurement, assessing respondents’ own perceptions about their resilience. In further studies, assessment of resilience based on diverse criteria such as developmental outcome, process, functionality etc., or observational procedures could address bias in measuring resilience. Thirdly, the cross-sectional nature of this study cannot support any causal relationship between variables. In order to draw firm cause and effect relationships, further research might utilize experimental designs.

Conclusion

A recent perspective shift has emphasized multilevel and interactional understanding of resilience. This study provides information about interactions among different protective factors in resilience of adolescents. Further research should continue to investigate how proximal and distal protective agents function in youth resilience. Understanding complex interactions between protective and risk factors may also provide a higher level of understanding regarding adolescent resilience.

Footnotes

Authors’ Note

This study was conducted under my dissertation process. Prof. Ayhan Demir, the second author of this article, was also the supervisor of my dissertation.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

Özden Sevil-Gülen

Author Biographies

Özden Sevil-Gülen is a psychological counselor in Mono Personal Development, Education and Counseling Center in İzmir, Turkey. Her main research interests concern resilience, psychological trauma, crisis counseling and social emotional learning.

Ayhan Demir is a professor in Department of Educational Sciences at Middle East Technical University in Ankara, Turkey. His research focuses on loneliness, social skills, conflict resolution, assertiveness and procrastination.

References

Atik

E. L.

(2013). The role of self-reflection and insight in the relationship between high school adolescents’ attachment styles and psychological resilience degree (Unpublished master’s thesis). Istanbul Bilim University.

Aydın

Sarı

S. V.

Şahin

(2014). Parental acceptance/involvement, self-esteem and academic achievement: The role of hope as a mediator. Cumhuriyet International Journal of Education, 3(4), 37–48.

Aydın-Sünbül

(2016). The relationship between mindfulness and resilience among adolescents: Mediating role of self-compassion and difficulties in emotion regulation (Unpublished doctoral dissertation). Middle East Technical University.

Barber

B. K.

(1997). Introduction: Adolescent socialization in context-the role of connection, regulation, and autonomy in the family. Journal of Adolescent Research, 12(1), 5–11.

Baumeister

R. F.

Leary

M. R.

(1995). The need to belong: Desire for interpersonal attachments as a fundamental human motivation. Psychological Bulletin, 117(3), 497–529.

Baumrind

(1991). The influence of parenting style on adolescent competence and substance use. Journal of Early Adolescence, 11(1), 56–95.

Benard

(2004). Resiliency: What we have learned? WestEd.

Bentler

P. M.

(1990). Comparative fit indexes in structural models. Psychology Bulletin, 107(2), 238–246.

Bronfenbrenner

(1979). Ecology of human development. Harvard University Press.

10.

Brown

A. T.

(2015). Confirmatory factor analysis for applied research (2nd ed.). The Guilford Press.

11.

Browne

M. W.

Cudeck

(1993). Alternative ways of assessing model fit. In Bollen

K. A.

Long

J. S.

(Eds.), Testing structural equation models (pp. 136–162). Sage Publications.

12.

Cauce

A. M.

Stewart

A. J.

Rodriguez

M. M. D.

Cochran

B. N.

Ginzler

(2013). Overcoming the odds: Adolescent development in the context of urban poverty. In Luthar

S. S.

(Ed.), Resilience and vulnerability: Adaptation in the context of childhood adversities (pp. 343–463). Cambridge University Press.

13.

Chiu

M. M.

Chow

B. W. Y.

McBride

Mol

S. T.

(2016). Students’ sense of belonging at school in 41 countries: Cross-cultural variability. Journal of Cross-Cultural Psychology, 47(2), 175–196.

14.

Cobb

(1976). Social support as a moderator of life stress. Psychosomatic Medicine, 38(5), 300–314.

15.

Cohen

Wills

T. A.

(1985). Stress, social support, and the buffering hypothesis. Psychological Bulletin, 98(2), 310–357.

16.

Colarossi

L. G.

Eccles

J. S.

(2003). Differential effects of support providers on adolescents’ mental health. Social Work Research, 27(1), 19–30.

17.

Conger

R. D.

Conger

K. J.

Elder

G. H.

Jr. (1997). Family economic hardship and adolescent adjustment: Mediating and moderating processes. In Duncan

G. J.

Brooks-Gunn

(Eds.), Consequences of growing up poor (pp. 288–310). Russell Sage Foundation.

18.

Çuhadaroğlu

(1985). Self-esteem in adolescents (Unpublished doctoral dissertation). Hacettepe University.

19.

Dayıoğlu

(2008). Resilience in university entrance examination applicants: The role of learned resourcefulness, perceived social support, and gender (Unpublished master’s thesis). Middle East Technical University.

20.

Dias

P. A.

Cadime

(2017). Protective factors and resilience in adolescents: The mediating role of self-regulation. Psicología Educativa, 23(1), 37–43.

21.

Dubow

E. F.

Ullman

D. G.

(1989). Assessing social support in elementary school children: The survey of children’s social support. Journal of Clinical Child Psychology, 18(1), 52–64.

22.

Eccles

J. S.

Midgley

Wigneld

Buchanan

C. M.

Reuman

Flanagan

Maclver

(1993). Development during adolescence: The impact of stage environment on young adolescents’ experiences in schools and in families. American Psychologist, 48(2), 90–101.

23.

Fletcher

Sarkar

(2013). Psychological resilience: A review and critique of definitions, concepts and theory. European Psychologist, 18(1), 12–23.

24.

Gökler

(2007). The Turkish adaptation study of social support appraisals scale to be used with children and adolescents: Factor structure, validity and reliability. Turkish Journal of Child and Adolescent Mental Health, 14(2), 90–99.

25.

Goodenow

(1993). The psychological sense of school membership among adolescents: Scale development and educational correlates. Psychology in the Schools, 30(1), 79–90.

26.

Gündaş

(2013). The role of self-construal and attachment styles as the predictors of psychological resilience of high school students (Unpublished master’s thesis). Gaziosmanpaşa University.

27.

Harter

(2006). The development of self-esteem. In Kernis

M. H.

(Ed.), Self-esteem issues and answers: A sourcebook of current perspectives (pp. 144–150). Psychology Press.

28.

Helsen

Vollebergh

Meeus

(2000). Social support from parents and friends and emotional problems in adolescence. Journal of Youth and Adolescence, 29(3), 319–340.

29.

Bentler

P. M.

(1999). Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural Equation Modeling: A Multidisciplinary Journal, 6(1), 1–55.

30.

Jöreskog

Sörbom

(1993). Structural equation modeling with SIMPLIS command language. Lawrence Erlbaum Associates.

31.

Kia-Keating

Ellis

B. H.

(2007). Belonging and connection to school in resettlement: Young refugees, school belonging, and psychosocial adjustment. Clinical Child Psychology and Psychiatry, 12(1), 29–43.

32.

Kline

B. R.

(2016). Principles and practice of structural equation modeling (4th ed.). Guilford Press.

33.

Kumpfer

L. K.

(1999). Factors and processes contributing to resilience: The resilience framework. In Glantz

M. D.

Johnson

J. L.

(Eds.), Resilience and development: Positive life adaptations (pp. 179–224). Kluwer Academic/Plenum Publishers.

34.

Lamborn

Mounts

Steinberg

Dornbusch

(1991). Patterns of competence and adjustment among adolescents from authoritative, authoritarian, indulgent and neglectful homes. Child Development, 62(5), 1049–1065.

35.

Lazarus

R. S.

Folkman

(1984). Coping and adaptation. In Gentry

W. D.

(Ed.), The handbook of behavioral medicine (pp. 282–325). Guilford Press.

36.

Mardia

K. V.

(1975). Assessment of multinormality and the robustness of Hotelling’s T2 test. Applied Statistics, 24(2), 163–171.

37.

Masten

A. S.

Best

K. M.

Garmezy

(1990). Resilience and development: Contributions from the study of children who overcome adversity. Development and Psychopathology, 2(4), 425–444.

38.

Masten

A. S.

Reed

M. J.

(2002). Resilience in development. In Snyder

C. R.

Lopez

S. J.

(Eds.), Handbook of positive psychology (pp. 74–88). Oxford University Press.

39.

Matsugana

(2008). Item parceling in structural equation modeling: A primer. Communication Methods and Measures, 2(4), 260–293.

40.

Nuttman-Shwartz

(2018). The moderating role of resilience resources and sense of belonging to the school among children and adolescents in continuous traumatic stress situations. The Journal of Early Adolescence, 39(1), 1–25.

41.

Olsson

C. A.

Bond

Burns

J. M.

Vella-Brodrick

D. A.

Sawyer

S. M.

(2003). Adolescent resilience: A concept analysis. Journal of Adolescence, 26(1), 1–11.

42.

Onat

(2010). The effects of parental attitudes-which are perceived as democratic and authoritarian-on first class of high school students? Resiliency level investigated (Unpublished master’s thesis). Maltepe University.

43.

Prince-Embury

P. S.

Saklofske

H. D.

(2014). Building a science of resilience intervention for youth. In Prince-Embury

P. S.

Saklofske

H. D.

(Eds.), Resilience interventions for youth in diverse populations (pp. 3–12). Springer.

44.

Rosenberg

(1965). Society and adolescence self-image. Princeton University Press.

45.

Sagone

De Caroli

M. E.

(2014). Relationships between psychological wellbeing and resilience in middle and late adolescents. Procedia - Social and Behavioral Sciences, 141, 881–887.

46.

Sarı

(2013). Lise öğrencilerinde okula aidiyet duygusu [Sense of school belonging among high school students]. Anadolu Üniversitesi Sosyal Bilimler Dergisi, 13(1), 147–160.

47.

Schumacker

R. E.

Lomax

R. G.

(1996). A beginner’s guide to structural equation modeling. Lawrence Erlbaum Associates.

48.

Seidman

Peterson

(2003). Holistic contextual perspectives on risk, protection and competence among low-income urban adolescents. In Luthar

S. S.

(Ed.), Resilience and vulnerability: Adaptation in the context of childhood adversities (pp. 318–342). Cambridge University Press.

49.

Şeker

(2011). Research of quality of life in İstanbul. İstanbul Chamber of Commerce Publications.

50.

Steinberg

Lamborn

S. D.

Darling

Mounts

N. S.

Dornbusch

S. M.

(1994). Over- time changes in adjustment and competence among adolescents from authoritative, authoritarian, indulgent, and neglectful families. Child Development, 65(3), 754–770.

51.

Steinberg

Vandell

D. L.

Bornstein

M. H.

(2011). Development: Infancy through adolescence. Cengage Learning.

52.

Stepleman

M. L.

Wright

E. D.

Bottonari

A. K.

(2009). Socioeconomic status: Risks and resilience. In Loue

Sajatovic

(Eds.), Determinants of minority mental health and wellness (pp. 273–302). Springer.

53.

Strudwicke

(2000). Sense of belonging and self-esteem: What are the implications for educational outcomes of secondary school students: A literature review (Unpublished master’s thesis). Edith Cowan University, Australia.

54.

Tabachnick

B. G.

Fidell

L. S.

(2013). Using multivariate statistics (6th ed.). Pearson.

55.

Türkiye İstatistik Kurumu (TÜİK). (2013). Seçilmiş göstergelerle İstanbul 2013. Türkiye İstatistik Kurumu Matbaası.

56.

Uçanok

Güre

(2012). Perceived economic strain and psychological well-being: The mediational role of parental relations in Turkish early adolescents. Journal of Early Adolescence, 34(6), 685–711.

57.

Ungar

(2012). Social ecologies and their contribution to resilience. In Ungar

(Ed.), The social ecology of resilience: A handbook of theory and practice (pp. 13–32). Springer.

58.

van Harmelen

A. L.

Kievit

R. A.

Ioannidis

Neufeld

Jones

P. B.

Bullmore

Goodyer

. (2017). Adolescent friendships predict later resilient functioning across psychosocial domains in a healthy community cohort. Psychological Medicine, 47(13), 1–11.

59.

van Ryzin

M. J.

Gravely

A. A.

Roseth

C. J

. (2009). Autonomy, belongingness, and engagement in school as contributors to adolescent psychological wellbeing. Journal of Youth and Adolescence, 38(1), 1–12.

60.

Veselska

Geckova

A. M.

Gajdosova

Orosova

van Dijk

J. P.

Reijneveld

S. A.

(2010). Socioeconomic differences in self-esteem of adolescents influenced by personality, mental health and social support. European Journal of Public Health, 20(6), 647–652.

61.

Wagnild

G. M.

(2010). The resilience scale user’s guide for the US English version of the resilience scale and the 14-item resilience scale (RS-14). The Resilience Center.

62.

Wagnild

G. M.

Young

H. M.

(1993). Development and psychometric evaluation of the Resilience Scale. Journal of Nursing Measurement, 1(2), 165–178.

63.

Yılmaz

(2000). Relationships between marital adjustment, child perception of parenting style, and children’s, adolescents’, and young adults’ academic achievement and self-perceptions (Unpublished doctoral dissertation). Hacettepe University.

64.

Zakeri

Jowkar

Razmjoee

(2010). Parenting styles and resilience. Procedia Social and Behavioral Sciences, 5, 1067–1070.

65.

Zimmerman

M. A.

Brenner

A. B.

(2010). Resilience in adolescence: Overcoming neighborhood disadvantage. In Reich

J. W.

Zautra

A. J.

Hall

J. S.

(Eds.), Handbook of adult resilience (pp. 283–308). Guilford Press.