Minors constitute a marked proportion of all refugee entrants to Australia, with approximately 39% of offshore visas in 2013–14 and 15% of onshore visas in 2012–13 granted to refugees aged below 18 years (Department of Immigration and Border Protection, 2014). The majority of studies on minors from refugee backgrounds have found that they experience elevated levels of psychopathology, particularly depression, anxiety, and post-traumatic stress disorder (PTSD; de Anstiss, Ziaian, Procter, Warland, & Baghurst, Reference de Anstiss, Ziaian, Procter, Warland and Baghurst2009; Ehntholt & Yule, Reference Ehntholt and Yule2006). However, there has also been evidence of positive adaptation and development. High levels of functioning, adjustment, and coping have been observed in many young refugees (e.g., Arnetz, Rofa, Arnetz, Ventimiglia, & Jamil, Reference Arnetz, Rofa, Arnetz, Ventimiglia and Jamil2013), and longitudinal research with refugee youth has generally shown a considerable decrease in the prevalence of psychopathology over time (Almqvist & Broberg, Reference Almqvist and Broberg1999; Sack, Him, & Dickason, Reference Sack, Him and Dickason1999; Steel, Silove, Phan, & Bauman, Reference Steel, Silove, Phan and Bauman2002). As such, understanding the factors that might promote recovery, adjustment, and coping, and thereby provide conditions for greater wellbeing among refugee youth — so-called ‘protective factors’ — is an area worthy of research attention.
Wellbeing in Refugee Youth
The majority of research in the refugee field has been directed towards understanding wellbeing by measuring psychopathology, and positive and subjective dimensions of wellbeing have been investigated significantly less. This reflects a traditional prioritisation in psychology research of a deficit-based mental health model over one based on positive psychological functioning (Keyes, Reference Keyes2007; Ryff & Keyes, Reference Ryff and Keyes1995). There is emerging evidence that, rather than being on the same continuum, positive wellbeing is located on a separate dimension to mental illness, and separate or additional conditions are therefore considered necessary for an individual to flourish (Keyes, Reference Keyes2002, Reference Keyes2007). These conditions are considered to encompass subjective aspects, such as happiness, positive affect, and satisfaction with life (e.g., Diener, Reference Diener1984; Lyubomirsky & Lepper, Reference Lyubomirsky and Lepper1999), and positive psychological aspects, such as autonomy, personal development, self-acceptance, and life purpose (e.g., Ryff, Reference Ryff1989; Waterman, Reference Waterman1993). It was considered important to explore multiple dimensions of young refugees’ wellbeing in the current study, and wellbeing is therefore measured on two separate dimensions; first, as an absence of psychological distress; and second, as the presence of positive and subjective wellbeing aspects. It should be noted that these constructs are influenced by Western notions of wellbeing; operationalisations that attempt to assess wellbeing from the cultural standpoints of the study participants would be more appropriate but are outside the scope of this study. While there is some understanding of both the socio-demographic and psychosocial factors that protect against psychological distress among refugee youth, considerably less is known about the conditions that may account for positive wellbeing; understanding factors that influence positive wellbeing, in addition to psychopathology, is therefore a focus of this study.
Socio-Demographic Protective Factors
Socio-demographic variables that may act as protective factors among refugee youth include gender, age, education level, time passed since the trauma, and visa type (Fazel, Reed, Panter-Brick, & Stein, Reference Fazel, Reed, Panter-Brick and Stein2012; Reed, Fazel, Jones, Panter-Brick, & Stein, Reference Reed, Fazel, Jones, Panter-Brick and Stein2012). There is an emerging base of evidence, for example, that preadolescent and younger adolescent refugee children may experience fewer depressive symptoms than their older counterparts (Hasanović, Sinanović, & Pavlović, Reference Hasanović, Sinanović and Pavlović2005). Similarly, girls appear to have a higher prevalence of internalising disorders, such as depression, anxiety, and loneliness (Bean, Eurelings-Bontekoe, & Spinhoven, Reference Bean, Eurelings-Bontekoe and Spinhoven2007), whereas boys tend to display more externalising disorders, such as disruptive behaviours and oppositional trends (Mels, Derluyn, Broekaert, & Rosseel, Reference Mels, Derluyn, Broekaert and Rosseel2010). Gender may therefore predispose refugee youth to encounter their challenges in different ways. Higher levels of education achievement may prevent refugees from securing a job and can predict higher levels of psychopathology (Slodnjak, Kos, & Yule, Reference Slodnjak, Kos and Yule2002). An increase in the time since trauma appears to correlate with a decrease in symptomatology (Almqvist & Broberg, Reference Almqvist and Broberg1999). There is also significant evidence that undetermined refugee status is associated with higher levels of psychological distress, including depression and PTSD (Heptinstall, Sethna, & Taylor, Reference Heptinstall, Sethna and Taylor2004; Steel et al., Reference Steel, Momartin, Silove, Coello, Aroche and Tay2011). Conversely, the efficient resolution of asylum claims has been found to reduce distress in children from refugee backgrounds (Heptinstall et al., Reference Heptinstall, Sethna and Taylor2004).
Psychosocial Protective Factors
Psychosocial factors that promote wellbeing may be grouped as individual-level, family and community-level, and society-level factors. Examples of individual-level factors are personal resilience, the degree of acculturation into the host society's culture, and faith in a religion or spirituality (Fazel et al., Reference Fazel, Reed, Panter-Brick and Stein2012; Montgomery, Reference Montgomery2010; Reed et al., Reference Reed, Fazel, Jones, Panter-Brick and Stein2012). Family and community-level factors include family health and functioning, social support from peers and community members, and a sense of school connectedness (Fazel et al., Reference Fazel, Reed, Panter-Brick and Stein2012; Porter & Haslam, Reference Porter and Haslam2005; Reed et al., Reference Reed, Fazel, Jones, Panter-Brick and Stein2012). Access to health care and education, and the efficiency with which immigration processes are resolved are examples of factors at the societal level (Fazel et al., Reference Fazel, Reed, Panter-Brick and Stein2012; Reed et al., Reference Reed, Fazel, Jones, Panter-Brick and Stein2012; Montgomery, Reference Montgomery2010). The present study was concerned with examining the contribution of certain psychosocial factors on both psychopathology and positive wellbeing among refugee youth, while controlling for the effect of socio-demographic variables. The main focus was on factors that can be fostered and influenced within a school environment, and therefore school connectedness, acculturation, and resilience were chosen. The socio-demographic variables considered most applicable to the participant cohort and were therefore controlled for were age, gender, length of stay in Australia, years of schooling, and visa type.
School connectedness
School connectedness can be described as ‘the extent to which students feel personally accepted, respected, included, and supported by others in the school and classroom’ (Goodenow, Reference Goodenow1993, p. 81). Clear links between school connectedness and higher academic outcomes (e.g., Maurizi, Ceballo, Epstein-Ngo, & Cortina, Reference Maurizi, Ceballo, Epstein-Ngo and Cortina2013), improved self-esteem (e.g., Osterman, Reference Osterman2000), and lower levels of depression and anxiety (Shochet, Dadds, Ham, & Montague, Reference Shochet, Dadds, Ham and Montague2006) have been identified in many cohorts. In one study, school connectedness accounted for between 38% and 55% of the covariance with depression and between 9% and 16% of that with anxiety (Shochet et al., Reference Shochet, Dadds, Ham and Montague2006). As schools are one of the first and most influential systems that young refugees come into contact with after resettlement, it can be expected that school connectedness plays an important role in the wellbeing of displaced youth. (Kia-Keating & Ellis, Reference Kia-Keating and Ellis2007). In a study of Somali adolescents resettled in the United States, a greater sense of school connectedness was associated with reduced depression and increased self-efficacy, regardless of the level of past exposure to hardship (Kia-Keating & Ellis, Reference Kia-Keating and Ellis2007). Other studies have found that, among student refugees, an increased sense of school belonging protected against depression (Rousseau, Drapeau, & Platt, Reference Rousseau, Drapeau and Platt2004) and anxiety (Sujoldzić Peternel, Kulenović, & Terzić, Reference Sujoldzić, Peternel, Kulenović and Terzić2006), and was positively linked to self-esteem (Sujoldzić et al., Reference Sujoldzić, Peternel, Kulenović and Terzić2006).
Acculturation
Acculturation is a process of cultural and psychological change that involves integrating ideas, beliefs, values, and behaviours from one's home culture into life in a newly adopted culture (Berry, Reference Berry2005). The acculturation strategy of integration — retaining some aspects of the original culture and adopting dimensions of the new culture — tends to lead to the most effective psychological adjustment across diverse population samples (Sam & Berry, Reference Sam and Berry2010; e.g., Kovacev & Shute, Reference Kovacev and Shute2004). Alternatively, either retaining the individual's own culture or assimilating entirely to the values of the host society is not predictive of psychosocial adjustment among young refugees (Kovacev & Shute, Reference Kovacev and Shute2004). Problems with the process of adopting, modifying, and relinquishing cultural values can lead to acculturative stress, which is an important risk factor for psychological distress (Lincoln, Lazarevic, White, & Ellis, Reference Lincoln, Lazarevic, White and Ellis2016). Acculturative stress may be exacerbated by factors in the host society, including adverse asylum processes and discrimination (Brough, Gorman, Ramirez, & Westoby, Reference Brough, Gorman, Ramirez and Westoby2003; Phillimore, Reference Phillimore2011), and individual-level variables, such as lack of resilience, low socioeconomic status, and poor proficiency in the host country's language (Brough et al., Reference Brough, Gorman, Ramirez and Westoby2003; Phillimore, Reference Phillimore2011).
Resilience
Resilience is defined as a modifiable dispositional capacity for appraising adversity positively that confers resistance against the development of mental illness (Arnetz et al., Reference Arnetz, Rofa, Arnetz, Ventimiglia and Jamil2013; Edward & Warelow, Reference Edward and Warelow2005). Few studies of resilience among refugees have directly measured the protective nature of dispositional resilience. An exception is a study by Arnetz et al. (Reference Arnetz, Rofa, Arnetz, Ventimiglia and Jamil2013), who found that resilience was a significant inverse predictor of general psychological distress, but not of PTSD symptoms, among Iraqi refugees in the United States (Arnetz et al., Reference Arnetz, Rofa, Arnetz, Ventimiglia and Jamil2013). In addition, related and possibly overlapping constructs, such as coping style and a sense of control (Hooberman, Rosenfeld, Rasmussen, & Keller, Reference Hooberman, Rosenfeld, Rasmussen and Keller2010), may ameliorate wellbeing among refugees.
The Present Study
Setting
The present study was undertaken at Milpera State High School (MSHS) in Brisbane. The school offers focused English language education and other core subjects to prepare newly arrived CALD migrants and refugees for entry to mainstream education. The majority of students attend the school for between 3 and 12 months. At the time of the study, there were over 24 nationalities represented in total. In addition to language tuition, the school assists students with the transition to living in Australia through two-way cultural education programs and instrumental support. Those students from refugee backgrounds who are identified as in need of therapeutic support can also access school-based intervention programs that use music and art therapy techniques.
Research Aims and Hypotheses
The aim of the present study was to explore the importance of psychosocial variables as contributors to two main constructs of wellbeing — an absence of psychopathology and a presence of positive wellbeing — in a culturally diverse sample of refugee children and adolescents. There were two hypotheses. First, it was hypothesised that higher levels of school connectedness, acculturation, and resilience would predict lower levels of depression and anxiety. Second, it was hypothesised that higher levels of school connectedness, acculturation, and resilience would be associated with higher levels of positive wellbeing.
Method
Participants
The sample consisted of 93 refugee students. There were 50 females (54% of participants) and 43 males (46% of participants). Overall, the mean age of participants was 15.46 (SD = 1.55) years (range = 12–18). Sixty-one percent of participants were on permanent visas and the remaining 39% had been granted bridging or temporary visas. Participants’ average length of stay in Australia was 10.24 months (SD = 5.24; range = 0–36). Twenty-four different birth countries and 32 different language groups were represented in the sample, the largest cohort being from Iran (22.6%). Other birth countries accounted for between 1.1% and 10.8% of the sample. Ninety-two percent of participants stated that they had a religion. Participants were from a diverse range of religions and denominations.
Predictor Measures
School connectedness
School connectedness was assessed through the Psychological Sense of School Membership (PSSM; Goodenow, Reference Goodenow1993). The PSSM is an 18-item self-report questionnaire that measures a student's sense of school belonging through a 5-point Likert scale ranging from 1 (not at all true) to 5 (completely true). Higher scores identify participants as more connected to their school. Test–retest reliability of the scale was good (r = .88) and internal consistency was adequate, with Cronbach's alphas ranging from .77 to .80 (Goodenow, Reference Goodenow1993). PSSM scores correlate significantly with social standing and student effort as rated by teachers, and with student-reported motivation and grades (Goodenow, Reference Goodenow1993).
Acculturation
Acculturation was measured through the acculturation subscale of the Adult Acculturation and Resiliency Scale (AARS; Khawaja, Moisuc, & Ramirez, Reference Khawaja, Moisuc and Ramirez2014). The AARS is a 27-item scale with a stable three-factor structure of resiliency, acculturation, and spirituality. It aims to measure the positive development of adaptation and resilience (Khawaja et al., Reference Khawaja, Moisuc and Ramirez2014). Acculturation, which is perceived within the AARS as the successful management of the new and original culture, is tested through 11 items. Psychometric properties were assessed using a culturally and linguistically diverse (CALD) community sample. Internal consistency (α = .83) and test–retest reliability (r = .65) of the acculturation factor were satisfactory (Khawaja et al., Reference Khawaja, Moisuc and Ramirez2014). For the present study, scale items were minimally reworded to be more suitable for young participants. The Likert-scale range was reduced to three options: 1 (not at all), 2 (a little) and 3 (a lot). Higher scores indicated higher levels of acculturation.
Resilience
Resilience was tested through 8 items from the 14-item resilience subscale of the AARS (Khawaja et al., Reference Khawaja, Moisuc and Ramirez2014) in combination with 2 items from the Resilience Questionnaire for Middle-Adolescents in Township Schools (R-MATS; Mampane, Reference Mampane2012). The Likert-scale range was limited to three options, from 1 (not at all) to 3 (a lot). Higher scores indicated higher levels of resilience.
Resilience is perceived within the AARS as an individual's ability to effectively cope with unfavourable situations by drawing on interpersonal and intrapersonal strengths (Khawaja et al., Reference Khawaja, Moisuc and Ramirez2014). On the sample of CALD community members described above, the resilience subscale of the AARS was demonstrated to have adequate internal consistency (α = .80) and test–retest reliability (r = .89).
Items 12 and 19 from the R-MATS were added to the resilience scale to investigate an additional construct: participants’ perseverance in the face of adversity. The R-MATS has a stable four-factor structure of confidence and internal locus of control, social support, toughness and commitment, and achievement orientation (Mampane, Reference Mampane2012). For the present study, two items from the toughness and commitment factor were used. Exploratory analysis has shown the R-MATS to be internally consistent, with a Cronbach's alpha of 0.82 (Mampane, Reference Mampane2012).
Outcome Measures
Depression and anxiety
Depression and anxiety were tested via the two subscales of a brief 25-item version of the Hopkins Symptom Checklist (HSCL), which was specifically adapted for use with refugee populations (Mollica, Wyshak, de Marneffe, Khuon, & Lavelle, Reference Mollica, Wyshak, de Marneffe, Khuon and Lavelle1987). The psychometric properties of the HSCL-25 are robust (Hollifield et al., Reference Hollifield, Warner, Lian, Krakow, Jenkins, Kesler and Westermeyer2002). In refugee samples, the English version of the HSCL-25 was found to have good reliability (α = .86 for anxiety; α = .85 for depression) and adequate item-total correlations (r = .41–.67 for anxiety; r = .30–.61 for depression; Kleijn, Hovens, & Rodenburg, Reference Kleijn, Hovens and Rodenburg2001). For use in the present study, some items of the HSCL-25 were reworded slightly. For example, ‘Feel tense or keyed up’ was changed to ‘Feel tense’. The original four categories of response (not at all, a little, quite a bit and extremely) were retained, with higher scores indicating higher levels of symptomatology.
Positive wellbeing
Positive wellbeing was assessed via the Stirling Children's Wellbeing Scale (SCWBS; Liddle & Carter, n.d.), which measures emotional and psychological wellbeing in children aged from 8 to 15 years. Internal consistency (α = .83) and test–retest reliability (r = .75) of the scale are adequate (Liddle & Carter, n.d.). The scale's construct validity has been demonstrated through strong significant correlations between the SCWBS and both the Warwick-Edinburgh Mental Wellbeing Scale and the DuBois Self-Esteem Scale. For the present study, the Likert-scale range was limited to three levels, from 1 (not at all) to 3 (a lot), with higher scores representing higher levels of positive wellbeing.
Procedure
The current study formed part of a larger project at MSHS investigating refugee and migrant youth experiences. Ethics and health and safety clearances were obtained from the university and the respective education department. Information about the study and consent forms were disseminated to prospective participants by the school's guidance officer. Parent's and guardian's consent was required for students under 18 years. Before initiating test administration, the research team participated in student classes over a 4-week period to establish rapport and familiarise students to their presence at the school.
The students who had consented to the study were administered the battery in one to two individual sessions of approximately 45 minutes at MSHS. Data pertaining to visa type and date of arrival was obtained from school enrolment records and verified with the participant during the interview. All questions were read in English by the researcher. Telephone interpreters were engaged to assist with the data collection using the participant's preferred language. At the completion of the interview, participants were debriefed and thanked for their involvement.
Results
Preliminary Analyses
A sample size of 91 young people was calculated as being necessary for a power of .80; based on an expected medium effect size of approximately .15 and using a significance criterion of .05 (Cohen, Reference Cohen1992). The data were screened for accuracy and the internal consistency of the scales and subscales was examined. Cronbach's alphas of the school connectedness, acculturation, resilience, and positive wellbeing scales were α = .87, α = .75, α = .85, and α = .93 respectively. Cronbach's alphas of the HSCL depression and anxiety subscales were α = .86 and α = .78 respectively.
Bivariate correlation analyses, as presented in Table 1, revealed that all psychosocial variables (school connectedness, acculturation, and resilience) were significantly negatively correlated with depressive and anxious symptoms. A greater length of stay in Australia and fewer years of schooling prior to arrival in Australia were significantly associated with higher levels of depression. All psychosocial variables were significantly positively correlated with positive wellbeing. Additionally, having a permanent visa was significantly related to higher levels of wellbeing.
TABLE 1 Pearson's Correlations of Depression, Anxiety and Positive Wellbeing
Note: N = 93.
*p < .05, 1-tailed. **p < .01, 1-tailed.
Multiple Regression Analyses
Hierarchical multiple regression analyses (MRAs) were performed to estimate the proportion of variance in wellbeing that can be accounted for by school connectedness, resilience, and acculturation, while controlling for the sociodemographic factors of gender, age, length of stay, years of schooling, and visa type. For each indicator of wellbeing (depression, anxiety, and positive wellbeing) a separate MRA was conducted to test the hypothesis that the protective factors at the focus of this study influence the wellbeing of young refugees. In each MRA, the sociodemographic variables were entered on step 1, and the psychosocial variables were entered together on step 2. Relevant assumptions were first tested and it was revealed that there were significant departures from normality in the sample distributions for all predictor and outcome variables. Bootstrapping procedures using sampling distributions of 1,000 were therefore used in the MRAs, hence bootstrapped values for the confidence intervals of unstandardised coefficients are reported. Additionally, there were departures from linearity and homoscedasticity of residuals. These assumptions are discussed further in relation to their respective outcome variables. Other assumptions were met, and testing indicated that outliers did not exert an undue influence on results.
Depression
As visual inspection of residual scatterplots revealed a marginal breach of homoscedasticity, it was controlled for by performing MRA through PROCESS using heteroscedastic consistent standard errors. Overall, the behaviour of variables and their relationships with one another did not differ markedly between the original multiple regression and PROCESS analyses, and therefore results of the original MRA are presented.
On step 1 of the hierarchical MRA, age, gender, length of stay, years of schooling, and visa type accounted for a non-significant 14.2% of the variance in depression, R 2 = .14, F(5, 50) = 1.65, p = .163. On step 2, school connectedness, acculturation, and resilience were added to the model and accounted for an additional 33.1% of the variance in depression, R 2change = .33, F change(3, 47) = 9.83, p < .001. In combination, all predictor variables explained 47.3% of the variance in depression, R 2 = .47, adjusted R 2 = .38, F(8, 47) = 5.27, p < .001. According to Cohen's (Reference Cohen1988) conventions, this represents a large effect. Unstandardised regression coefficients (B) with bootstrapped confidence intervals, standardised regression coefficients (β), and squared semi-partial correlations (sr2) for each predictor in the regression model are presented in Table 2. Table 2 shows that school connectedness and acculturation were significant predictors of depression (sr2 = .04 and .05 respectively) in the final regression model after accounting for the effects of sociodemographic factors. These represent small effect sizes according to Cohen (1998).
TABLE 2 Unstandardised (B) and Standardised (β) Regression Coefficients and Squared Semi-Partial Correlations (sr 2) for Each Predictor in a Hierarchical Regression Model Predicting Depression Among Refugee Youth
Note: N = 93. CI = confidence interval.
*p < .05; **p < .01.
Anxiety
As visual inspection of residual scatterplots revealed that the homoscedasticity assumption had been breached, MRA was performed through PROCESS using heteroscedastic consistent standard errors. As coefficient values of the variables and the relationships between variables did not differ markedly between the original multiple regression and PROCESS analyses, results of the original MRA are presented.
On step 1 of the hierarchical MRA, gender, age, length of stay, years of schooling, and visa type accounted for a non-significant 8.6% of the variance in anxiety scores, R 2 = .09, F(5, 50) = 0.94, p = .461. On step 2, 31.4% of the variance in anxiety was accounted for by the addition of school connectedness, acculturation, and resilience, R 2change = .31, F change(3, 47) = 8.22, p < .001. Together, all predictor variables explained 40.1% of the variance in anxiety, R 2 = .40, adjusted R 2 = .30, F(8, 47) = 3.93, p < .001. According to Cohen's (Reference Cohen1988) conventions, an effect of this size can be considered ‘large’. Table 3 presents unstandardised regression coefficients (B) with bootstrapped confidence intervals, standardised regression coefficients (β), and squared semi-partial correlations (sr2) for each predictor in the regression model. As can be seen from Table 3, acculturation was a significant predictor of anxiety in the final regression model (sr2 = .09), representing a medium effect by Cohen's (Reference Cohen1988) conventions.
TABLE 3 Unstandardised (B) and Standardised (β) Regression Coefficients and Squared Semi-Partial Correlations (sr 2) for Each Predictor in a Hierarchical Regression Model Predicting Anxiety Among Refugee Youth
Note: N = 93. CI = confidence interval.
*p < .05; **p < .01.
Positive wellbeing
Results of the initial hierarchical MRA showed that, on step 1, age, gender, length of stay, years of schooling, and visa type accounted for a significant 23.8% of the variance in positive wellbeing, R 2 = .24, F (5, 50) = 3.13, p = .016. When school connectedness, acculturation, and resilience were added to the regression equation on step 2, those variables together accounted for an additional 55% of the variance in positive wellbeing, R 2change = .55, F change(3, 47) = 40.79, p < .001. In combination, all predictor variables explained 78.9% of the variance in positive wellbeing, R 2 = .79, adjusted R 2 = .75, F(8, 47) = 21.92, p < .001. By Cohen's (Reference Cohen1988) conventions, this represents a large effect. Unstandardised regression coefficients (B) with bootstrapped confidence intervals, standardised regression coefficients (β), and squared semi-partial correlations (sr2) for each predictor in the regression model are presented in Table 4. Table 4 shows that school connectedness (sr2 = .10), acculturation (sr2 = .02), and resilience (sr2 = .06) emerged as significant psychosocial predictors of positive wellbeing. In addition, visa type was a sociodemographic variable that presented as a significant predictor of positive wellbeing (sr2 = .03). These effects are of small to medium magnitudes (Cohen, Reference Cohen1988).
TABLE 4 Unstandardised (B) and Standardised (β) Regression Coefficients and Squared Semi-Partial Correlations (sr 2) for Each Predictor in a Hierarchical Regression Model Predicting Positive Sense of Wellbeing Among Refugee Youth
Note: N = 93. CI = confidence interval.
*p < .05. **p < .01.
Discussion
The purpose of the study was to investigate the relationship between psychosocial protective factors and multiple dimensions of wellbeing while controlling for socio-demographic variables in a culturally diverse sample of young refugees. The hypotheses were partially supported. School connectedness and acculturation were negatively associated with depression scores, while acculturation was negatively correlated with anxiety scores. Further, school connectedness, acculturation, and resilience were positively related with positive wellbeing and jointly accounted for a large effect. Visa type was the only sociodemographic variable that emerged as a significant predictor in the analyses, with results indicating that having a permanent visa was associated with higher levels of positive wellbeing.
Wellbeing Reflected by an Absence of Psychological Distress
The finding of a significant relationship between higher levels of school connectedness and lower levels of depression is well supported in other research among refugee youth and in the general population (e.g., Kia-Keating & Ellis, Reference Kia-Keating and Ellis2007). The findings accord with Fazel and colleagues’ (Reference Fazel, Reed, Panter-Brick and Stein2012) proposition that schools are uniquely placed to undertake vital aspects of care for refugee children and have an important role in mitigating long-term negative psychological sequelae by enhancing resilient behaviours, facilitating the development of peer relationships, acting as a stable social support system, and assisting integration. One particular implication of these findings may be further support for the development of interventions within school settings, with the specific aim of helping children overcome difficulties arising from forced migration (e.g., Tyrer & Fazel, Reference Tyrer and Fazel2014).
The significant correlation between higher levels of acculturation and lower levels of both depression and anxiety extends on an emerging area of research that suggests that employing a particular acculturation strategy — integration — is associated with better psychological functioning. Few other studies have examined the relationship between integration and depressive and anxious pathology in particular, focusing instead on measures such as self-worth, social support, and loneliness (e.g., Kovacev & Shute, Reference Kovacev and Shute2004). An exception is a recent study by Lincoln et al. (Reference Lincoln, Lazarevic, White and Ellis2016), which also found that, as compared to marginalised or separated refugee groups, lower depression levels were associated with an acculturative style that included participation in the host and original cultures.
It is particularly interesting that in the present study, acculturation was the only psychosocial variable to have a notable effect on anxiety scores, as this represents a previously unidentified relationship. The findings may be explained by the fact that successful acculturation resolves acculturative stress by reducing conflict between learning new aspects of the host culture and relinquishing aspects of the original culture (Berry, Reference Berry2005). The results suggest that integration may be a form of acculturation that resolves acculturative stress and conflict effectively, as asserted by Berry (Reference Berry2005). Furthermore, the results suggest that better mental health outcomes for refugees will ensue from implementing strategies that minimise acculturative stress, such as by facilitating young refugees to develop peer and community networks (Tousignant et al., Reference Tousignant, Habimana, Biron, Malo, Sidoli-LeBlanc and Bendris1999), and have positive experiences of multiculturalism (Brough et al., Reference Brough, Gorman, Ramirez and Westoby2003), and by developing programs that support young refugees to retain a balance between aspects of their home culture and dimensions of the new culture.
Positive Wellbeing
Psychosocial factors predicted a large proportion of the variance in positive wellbeing, with school connectedness and resilience being the strongest predictors. The outcomes regarding school connectedness in particular are supportive of prior research that found that refugee adolescents’ greater sense of belonging, particularly perceptions of status and inclusion, were strongly associated with greater subjective wellbeing after the first 8 years of resettlement (Correa-Velez, Gifford, & McMichael, Reference Correa-Velez, Gifford and McMichael2015). Factors that promote positive wellbeing in refugees are an understudied area, and there is little other research with which to compare the findings of the present study. However, the results are consistent with what is known about the foundations of positive wellbeing in the general population, which include supportive parenting and friendship networks, positive life experiences, and intimate attachments (Dodge, Daly, Huyton, & Sanders, Reference Dodge, Daly, Huyton and Sanders2012).
There is growing interest in the conditions that allow individuals to flourish, given emerging support for the notion that while the majority may be free of mental illness, they may not necessarily enjoy good mental health (e.g., Keyes, Reference Keyes2002, Reference Keyes2007). Good mental health in the mainstream population is believed to be associated with a range of positive outcomes. For example, among youth, subjective wellbeing is believed to mitigate the negative effects of stressful life experiences, protect against the development of psychological and behavioural problems, and be correlated with physical health (Reed et al., Reference Reed, Fazel, Jones, Panter-Brick and Stein2012). The findings of the present study are therefore important, as they indicate that fostering a sense of school belonging and building personal resilience, in particular, may help to minimise psychological, behavioural, physical, and other stress-induced difficulties among young refugees.
It is notable that permanent visa status emerged as a significant sociodemographic predictor of wellbeing, accounting for a small proportion of the variance in the outcome. The findings are aligned with majority of existing research, which shows strong negative effects of insecure asylum status on the refugees on TPVs (a former visa class that conferred temporary refugee status on an ongoing basis and often for prolonged periods; Fazel et al., Reference Fazel, Reed, Panter-Brick and Stein2012; Reed et al., Reference Reed, Fazel, Jones, Panter-Brick and Stein2012). The population sample for the present study consisted of permanent refugees, as well as asylum seekers on temporary visa types, which differs to the former TPVs, so it is difficult to compare results of the present study with those that have preceded it. Despite this difference, the fact that permanent visa status accounted for a small proportion of the variance in wellbeing scores corroborates with existing evidence that being entitled to stay in the host country on a permanent basis is psychologically beneficial. This may be because permanent residency provides certainty and security (Fazel et al., Reference Fazel, Reed, Panter-Brick and Stein2012), and bestows additional rights and service access that minimise living difficulties (Nickerson, Steel, Bryant, Brooks, & Silove, Reference Nickerson, Steel, Bryant, Brooks and Silove2011).
Research Limitations and Future Directions
There were several research limitations, and the findings should be taken with caution. The sample size was small, limiting statistical power to detect significant regression effects. It was also drawn from a specialised educational setting that aims to provide a tailored and supportive learning environment, which restricts generalisation of the findings to broader refugee populations. It is recommended that future studies use a larger and more random sampling procedure. Moreover, this field of research would benefit from a more longitudinal approach to assess the wellbeing of young refugees in multiple contexts and after a longer duration of stay in their new country. Although it was outside the scope of the present study, one of the authors is following participants through a longer-term project with the aim of understanding the young refugees’ experiences over time, and particularly in a mainstream schooling context.
It was also outside the scope of the study to control for all demographic variables that may have accounted for differences in wellbeing states within the sample. Specifically, the data collection process did not include a reliable measure of English language level or assess other factors that may impact on refugee wellbeing, such as unaccompanied arrival or prior detention. Failing to control for these factors may have exerted an undue influence on results.
Amendments to data collection procedures to allow for cultural and language differences, and the fact that the project was based on largely Western constructs, need to be taken into account in the interpretation of results. In particular, the psychometric properties of the scales may have been altered by making CALD- and youth-friendly amendments. Additionally, the meaning of certain items may have been lost in translation as a result of administering questionnaires in an interview format through interpreters. Furthermore, culturally specific understandings or definitions were not reflected in the constructs; rather, the constructs were situated in Western-based epistemologies and measured using questionnaires developed largely with Western populations. Qualitative methodology would be an appropriate means through which to better understand concepts that have culturally specific or culturally influenced meanings. Hollifield and colleagues (Reference Hollifield, Warner, Lian, Krakow, Jenkins, Kesler and Westermeyer2002) suggest that greater robustness in refugee-related research can be achieved through developing instruments within community populations using empirical approaches that combine both qualitative and quantitative methods. It is therefore also suggested that the present findings be further investigated through quantitative research that employs assessment instruments specifically developed or adapted for use with young refugees.
Implications
The present study has important implications for policy and practice in the fields of education, mental health, and broader refugee governance. First, these findings can be used to shape refugee education environments characterised by respect for cultural diversity, supportive staff-student relations, and healthy peer interactions, with a specific focus on acculturation and school connectedness strategies. Second, the findings can be used to guide the ongoing development of targeted mental health interventions for depression and anxiety that facilitate a sense of school belonging and integration in particular. Finally, at the broader governmental level, the findings should be used to develop policies aimed at strengthening protective factors to not only mitigate mental illness but also promote mental health.
Conclusion
In spite of the limitations, the study was the first of its kind to investigate correlates of positive wellbeing, and it extended existing research into factors promoting an absence of depression and anxiety in young refugees. The outcomes indicated, first, that school belongingness and acculturation, as manifested by integration, may be protective against depression and anxiety among young refugees. Second, the study found that being strongly connected to school, well acculturated and resilient are conditions that may promote a positive sense of wellbeing in young refugees.
