CĂ©sar Diasa, Paulo Cadime, Irene
El objetivo principal de este estudio fue explorar el papel mediador de la autorregulación en la relación entre los factores protectores y la resiliencia. La muestra se compone de 393 adolescentes de cursos de bachillerato. Los participantes fueron evaluados mediante el Healthy Kids Resilience Assessment para recoger información sobre cuatro factores protectores-entorno escolar, entorno familiar, entorno de la comunidad y entorno entre pares-y resiliencia. Dos dimensiones de la autorregulación-establecimiento de objetivos y control de los impulsos-se evaluaron utilizando el Short Self-Regulation Questionnaire. Los resultados indicaron que el ambiente familiar, de la comunidad y entre iguales predijo significativamente los niveles de resiliencia de los estudiantes. El entorno familiar fue el principal predictor de la resiliencia. El establecimiento de objetivos y la capacidad de control de impulsos también fueron predictores de la resiliencia, pero no se encontraron pruebas de un efecto mediador de las dimensiones de la autorregulación sobre la relación entre los factores de protección y resiliencia. Se discuten los resultados y se presentan implicaciones para la prevención desde una perspectiva del desarrollo.
The main goal of this study was to explore the mediating role of self-regulation in the relationship between protective factors and resilience. The sample was composed of 393 adolescents who attended secondary education. Participants were assessed using the Healthy Kids Resilience Assessment to collect information about four protective factors-school, home, community and peer environment-and resilience. Two dimensions of self-regulation-goal setting and impulse control-were assessed using the Short Self-Regulation Questionnaire. The results indicated that home, community and peer environment predicted significantly the levels of resilience of the students. Home environment was the main resilience predictor. Goal setting and impulse control abilities were also predictors of resilience but no evidence was found for a mediating effect of the self-regulation dimensions on the relationship between protective factors and resilience. Results are discussed and implications for prevention from a developmental perspective are presented.
Adaptation to adversity is a priority research topic due to its importance to mental health promotion and risk behaviour prevention in children and adolescents. The study of resilience, which can be broadly defined as the dynamic process through which positive outcomes are achieved in the context of adversity ( Masten, 2001, 2014 ), can provide educators and governments with information about the resources that can be mobilised to promote positive outcomes in adverse environments ( UNICEF, 2013).
The effect of risk and protective factors on the resilience levels of the individuals has been widely studied. According to Rutter (1985) , risk factors are all stressful life events, such as poverty, family breakdown, experience of some form of violence, emotional loss, disease, unemployment, wars, disasters, or other factors that theoretically increase the likelihood of the onset of a problem or maintain the problem. However, research has shown that mere exposure to risk factors, even in substantial number, does not necessarily translate into risk behaviours ( Dillon et al., 2007 ), given that most people are able to deal with risk quite satisfactorily ( Bonanno, Galea, Bucciarelli, & Vlahov, 2007 ). In fact, several studies actually tend to focus on the effects of the promotion of protective factors, shifting the attention from risks to actions and strategies that are effective in the promotion of positive psychological adjustment and development ( Masten, 2001; Rutter, 1985, 2012; Zolkoski & Bullock, 2012 ). Protective factors are the individual characteristics or contextual/external conditions that work as dynamic mechanisms that help children and young people to resist or to balance the risks to which they are exposed to ( Rutter, 1985, 2012 ). Personality traits, perseverance, determination, self-efficacy, creativity, coherence, and self-awareness are some internal attributes that are positively related to resilience (e.g., Affi & Macmillan, 2011; Smokowski, Reynolds, & Bezrucko, 1999; Waaktaar, Christie, Borge, & Turgersen, 2004 ). Family support, positive appraisals, and quality interaction with parents ( Affi & Macmillan, 2011; Carbonell et al., 2002; Smokowski et al., 1999; Vanderbilt-Adriance & Shaw, 2008 ), good interpersonal relations ( Carbonell et al., 2002; Sameroff & Rosenblum, 2006 ), teacher and school support ( Brooks, 2006; Smokowski et al., 1999 ), as well as community relationships and resources ( Davies, Thind, Chandler, & Tucker, 2011; Sameroff & Rosenblum, 2006 ) are some of the most studied external protective factors.
A personal competence that looks particularly relevant to the resilience levels of individuals is their self-regulation ability ( Dishion & Connell, 2006; Gardner, Dishion, & Connell, 2008 ). In a recent review, Dillon et al. (2007) point out a number of factors at the personal level that are strongly correlated with the development of resilience. These factors the ability to: (a) operate a scheme to acknowledge, interpret, and predict potentially dangerous situations that occur in the social context; (b) develop a set of goals and regulate his/her behaviour in order to pursue these goals over time and in different contexts; and (c) promote positive beliefs about his/her self-regulatory capacity, to put his/her decisions into practice by using adequate skills and solve al problems. Most of these abilities are essentially self-regulation abilities, a process in which individuals take an active role shaping their own destiny even in contexts characterized by a high psychosocial risk ( Dishion & Connell, 2006; Masten & Coatsworth, 1998 ). Despite this evidence, few studies explore the mediating role of self-regulation in the relationship between protective factors and resilience. Therefore, this study aims to explore the existence of this mediating effect in a sample of adolescents, in order to contribute to a broader perspective about the relationship between protective factors and resilience, considering family-, peer-, school-, and community-related factors.From Protective Factors to the Mediating Role of Self-regulation in Resilience
Although different definitions of self-regulation exist in the literature (for a review, see Martin & McLellan, 2008 ), most of them seem to share the idea that this is an ability that involves the capacity to maintain effort and orientation toward a desired objective, while controlling the immediate impulses that arise ( Martin & McLellan, 2008; Neal & Carey, 2005 ). Successful adaptation to challenges appears then to depend on how individuals manage their emotions, think constructively, regulate and direct their behaviour, control automatic impulses, and act on the environment to change or decrease the sources of stress ( Compas, Connor- Smith, Saltzman, Harding Thomsen, & Wadsworth, 2001 ). Therefore, the study of self-regulation can be particularly relevant to understanding the resilience levels of the individuals, if we take into ac the necessity of managing these personal challenges to construct an adjusted life pathway ( Buckner, Mezzacappa, & Beardslee, 2009; Gestsdottir & Lerner, 2008; Lengua, 2002 ). Therefore, self-regulation exerts an effect in the adaptation to adversity, given that it is a process in which individuals take an active role in the activation, monitoring, inhibition, and/or adaptation of their behaviour, emotions, and cognitive strategies to achieve the desired goals ( Gestsdottir & Lerner, 2008; Moilanen, 2007 ).
Some individual and environmental factors, either related with family, peers, or community, are also positively related to self-regulation. Despite the consensus about self-regulation as a stable characteristic throughout age ( Buckner et al., 2009; Shoda, Mishel, & Peake, 1990 ), research has shown that it tends to be higher in girls ( Buckner et al., 2009; GarcĂa del Castillo & Dias, 2009; McCabe & Brooks-Gunn, 2007 ).
Family environmentâ€“family structural conditions, but especially affect, parental responsiveness, and control strategies and the exposure to adjusted modelsâ€“has also an important impact on children's and adolescentsâ€™ self-regulation skills ( Choe, Olson, & Sameroff, 2013; Colman, Hardy, Albert, Raffaelli, & Crockett, 2006; Karreman, van Tuijl, van Aken, & Dekovic, 2006; Moilanen, Shaw, & Fitzpatrick, 2010 ). At social level, some studies indicate a negative correlation between self-regulation and peer rejection, problematic or antisocial behaviours with peers ( Tangney, Baumeister, & Boone, 2004; Trentacosta & Shaw, 2009 ), but a positive correlation with the involvement and inclusion in the community ( Stenseng, Belsky, Skalicka, & WichstrĂ¸m, 2014; Stillmand & Baumeister, 2013 ). At school level, self-regulation has been positively related to school achievement ( Boekaerts, 2005; Duckworth & Seligman, 2005; McClellan & Cameron, 2011; Zimmerman, 1990 ) and higher levels of self-regulation have been found in students from regular or non-vocational courses when compared to students who attend vocational courses ( Jonker, Elferink-Gemser, Toering, Lyons, & Visscher, 2010; Koning & Boekaerts, 2005 ).
To summarise, individual and environmental factors, related either with school, home, community, or the group of peers, contribute to the development both of self-regulation and resilience in adolescents. Self-regulation also plays a central role in the development of resilience, given its importance in the mobilisation of resources in adverse environments, by setting goals and controlling impulses. In order to contribute to a broader understanding of the dynamics between the protective factors and resilience in adolescence, the goals of this study were twofold:
To test the existence of age, gender, and course-related differences in protective factors (school, family, community, and peers environment), self-regulation, and resilience.
To investigate if self-regulation mediates the relationship between protective factors (school, family, community, and peer environment) and resilience.
The initial sample d 404 students, but 11 participants did not complete at least one measure and were therefore discarded from the study. Therefore, the final sample d 393 students in grades 10 ( n = 323, 188 girls), 11 (n = 49, 25 girls) and 12 (n = 21, 14 girls), from three high schools located in northern Portugal. Participants were aged between 14 and 21 years old ( M age = 16.16, SD = 1.05). The participants attended either vocational courses ( n = 218, 55.5%) or high school regular courses ( n = 175, 44.5%). Students from vocational courses were older than students from regular courses, t(383) = -9.55, p < .001 (see Table 1 ). No association was found between gender and course type, ? 2(1) = 0.924, p = .321.
Age and Gender Distribution for the Students Who Attended Vocational and Regular High School Courses ( N = 393).
|M (SD)||16.56 (1.13)||15.67 (0.68)|
Note. N/I = no information provided about gender.
Healthy Kids Resilience Assessment (HKRA; Constantine, Bernard, & Diaz, 1999 ). The adaptation of the HKRA (version 6.0) to the Portuguese population ( Martins, 2007 ) was used. HKRA assesses four types of protective factorsâ€“school environment, home environment, community environment and peer environmentâ€“as well as resilience traits of high school students. The four types of protective factors are assessed with 33 items (nine items for each of the three first subscales and six items for the peer environment subscale). Items assess the existence of caring relationships, high expectations, and meaningful opportunities for participation in school, home, and community and the existence of caring relationships and high expectation messages among peers. Resilience is assessed with 18 items that cover six aspects: cooperation and communication, self-efficacy, empathy, problem-solving competences, self-knowledge, and goals and aspirations. Items are rated on a 4-point scale and the scale scores are calculated by taking the total sum of the items and dividing it by the number of items in each subscale. Regarding the reliability of the adaptation for the Portuguese population, the internal consistency coefficients were high for the resilience (? = .84), school environment (? = .83), home environment (? = .75) and community environment (? = .82) subscales. The internal consistency of the items that composed the peer environment subscale was slightly lower (? = .66), which can be attributable to the low number of items of the subscale ( Martins, 2007).
Short Self-Regulation Questionnaire (SSRQ; Carey, Neal, & Collins, 2004 ). Self-regulation was assessed using the SSRQ, which is originally composed of 31 items scored on a 5-point Likert scale, ranging from 1 ( strongly disagree) to 5 (strongly agree). Previous research (Carey et al., 2004 ) indicated that the SSRQ was a one-dimension measure that represented overall self-regulation capacity. However, subsequent research ( Neal & Carey, 2005 ) found evidence for a two-factor structure that represented two self-regulation dimensions: impulse control and goal setting. The Portuguese version also found evidence for this two-factor structure ( GarcĂa del Castillo & Dias, 2009 ). The Portuguese version comprises 29 items, from which 15 belong to the goal setting subscale (? = .85) and 14 to the impulse control subscale (? = .82).
Socio-demographic questionnaire . An additional questionnaire was used to collect information about demographic variablesâ€“gender, age, school gradeâ€“and education pathwaysâ€“school and type of high school course attended (vocational vs. regular high school).Procedure
Informed consent from school boards and parents of all students who attended the classes that were d in the sample were gathered prior to data collection. All students who were authorized by their parents and who were present in the day of data collection participated in the study voluntarily. Data were anonymised and confidentiality was assured. Participants completed questionnaires in classroom groups and received no material compensation for their participation.Statistical Analyses
After screening the data for missing values, differences in the protective factors (school, home, community, and peer environment), self-regulation (goal setting and impulse control), and resilience were examined using a series of independent samples t-tests.
To test the hypothesis that self-regulation mediates the relationship between protective factors (predictors) and resilience levels in adolescents (outcome variable), we used the analytic strategy outlined by Baron and Kenny (1986) . According to this strategy, three conditions should be present to detect a mediation effect: (1) the predictor is correlated with the outcome variable and the mediator; (2) the mediator predicts the outcome variable; and (3) the association between the predictor and the criterion variable diminishes or disappears when the mediator variable is controlled in the analysis. Therefore, Pearson correlation coefficients and a set of hierarchical regression analyses were performed to test for mediation. The following steps were followed in two regression models (one for each self-regulation dimension): (a) resilience was regressed on the protective factors in a first step; (b) self-regulation dimensions (impulse control and goal setting) were added separately to each model in a second step.Results Missing Values
Only 278 (70.7%) of the participants provided complete data on the SSRQ and 302 (76.8%) provided complete data on the HKRA. Therefore, 29.3% of the participants had at least one missing value on SSRQ and 23.2% at least one missing value on the HKRA. Participants with missing data and participants with complete data did not differ in terms of age, gender, grade, and course type. Therefore, the missing values in the item scores were d by the mean score of each item. After data imputation, complete data were obtained for the 393 participants in all measures.Differences according to Gender and Course Type
Table 2 shows means and standard deviations for all four protective factors, the two self-regulation dimensions, and resilience, as well as differences in these variables according to gender and course type (regular or vocational).
Differences in Protective Factors, Self-regulation and Resilience according to Gender and Course Type.
|Variables||Full Sample |
|Gender (N = 388)||Course Type (N = 393)|
|t||p||Regular M (SD)||Vocational M (SD)||t||p|
|School environment||1.67 (0.60)||1.68 (0.57)||1.67 (0.62)||0.13||.898||1.58 (0.63)||1.74 (0.56)||?2.68||.008|
|Home environment||2.12 (0.48)||2.16 (0.48)||2.06 (0.49)||1.99||.048||2.18 (0.49)||2.07 (0.47)||2.39||.018|
|Community environment||1.86 (0.67)||1.90 (0.67)||1.81 (0.68)||1.38||.169||1.87 (0.67)||1.86 (0.68)||0.08||.934|
|Peers environment||1.98 (0.45)||2.03 (0.43)||1.91 (0.49)||2.69||.008||1.99 (0.45)||1.97 (0.46)||0.27||.791|
|Goal setting||56.40 (7.50)||56.99 (7.26)||55.68 (7.76)||1.70||.090||57.52 (6.96)||55.51 (7.81)||2.66||.008|
|Impulse control||49.32 (8.61)||50.56 (8.53)||47.62 (8.51)||3.35||.001||50.86 (8.79)||48.09 (8.28)||3.21||.001|
|Resilience||2.09 (0.43)||2.15 (0.42)||2.02 (0.43)||2.91||.004||2.17 (0.41)||2.02 (0.43)||3.32||.001|
Girls obtained higher scores for home and peer environment than boys. Girls also reported higher impulse control abilities and resilience levels than boys. No differences between boys and girls were found for school and community environment or goal setting abilities.
Regarding the course type, students from vocational courses had lower levels of self-regulation (both goal setting and impulse control abilities) and resilience than students from regular courses. Students from regular courses reported better home environment but students from vocational courses reported better school environment (see Table 2 ). No differences between students from regular and vocational courses were found for community and peer environment.Relationship between Age, Protective Factors, Self-regulation, and Resilience
Bivariate correlations between the study variables are presented in Table 3 . Age had a negative correlation with impulse control, indicating that older students had less impulse control than younger students. No other variable had statistically significant correlations with age.
Pearson Product Moment Correlation Coefficients between Protective Factors, Self-regulation Dimensions and Resilience.
|1. School environment||-|
|2. Home environment||.34***||-|
|3. Community environment||.44***||.37***||-|
|4. Peers environment||.46***||.37***||.35***||-|
|5. Goal setting||.12*||.28***||.16**||.12*||-|
|6. Impulse control||.01||.26***||.11*||.01||.49***||-|
* p < .05, ** p < .01, *** p < .001.
Resilience correlated positively with all four protective factorsâ€“school, home, community, and peer environmentâ€“, as well as with goal setting and impulse control. All these correlation coefficients were moderate, varying between .30 and .52. Regarding the relationship between resilience and the protective factors, the highest correlation coefficient was obtained between resilience and home environment, followed by community environment.
The first self-regulation dimension, goal setting, correlated positively with all four protective factors, although the magnitude of the coefficients was low. However, the second self-regulation dimensionâ€“impulse controlâ€“correlated significantly only with home and community environment. However, these correlation coefficients between the two self-regulation dimensions and the protective factors were generally low (see Table 3).The Mediating Role of Self-regulation in the Relationship between Protective Factors and Resilience
According to the criteria set forth by Baron and Kenny (1986) , for a mediation effect to be present, the predictor(s) should be related both with the outcome variable and with the possible mediator(s). As can be seen in Table 3 , all four protective factors (predictor variables) are significantly correlated with resilience levels (outcome variable) and with goal setting abilities of students (mediator 1). The protective factors home and community environment were also significantly correlated with the impulse control abilities of students (mediator 2).
The second criteria by Baron and Kenny (1986) established that the mediator(s) should predict the outcome variable. In this study, both goal setting (? = 0.43, p < .001) and impulse control (? = 0.30, p < .001) abilities were statistically significant predictors of resilience.
The final criteria for establishing a mediation effect indicates that the association between the predictor and the outcome variable should diminish or disappear when the mediator is controlled in the analysis. The results of the regression analyses to test for this condition of mediating effect are presented in Table 4 . In model 1, the four protective factors were ed as predictors of the resilience levels in a first step, and the goal setting abilities (mediator 1) were ed in a second step. Home, community, and peer environment were significant resilience predictors in the first step, but not school environment. Home environment was the predictor with the strongest weight on the resilience level of the adolescents (see Table 4 ). When goal setting abilities were added to the model, all three predictors remained significant (see Table 4 ). Therefore, goal setting abilities do not seem to mediate the relationship between resilience and home, community, and peer environment, but are an independent predictor of the resilience levels achieved by adolescents.
Regression Analysis for Testing the Mediational Effects of the Self-regulation Dimensions in the Relationship between Protective Factors and Resilience.
|Model/Step||Predictors||R 2||?R 2||B||SE(B)||?|
|Step 1||School environment||.38||-||0.06||0.03||0.09|
|Step 2||School environment||.46||.08***||0.06||0.03||0.08|
|Step 2||School environment||.42||.04***||0.07||0.03||0.10|
* p < .05, ** p < .01, *** p < .001.
In model 2, the four protective factors were ed as predictors of the resilience levels in a first step, and the impulse control abilities (mediator 2) were added in a second step. Similarly to model 1, home, community, and peer environment were statistically significant resilience predictors in the first step. With the addition of impulse control abilities in the second step, all three predictors remained significant (see Table 4 ). These results also indicate that impulse control abilities do not mediate the relationship between protective factors and resilience, but are themselves an independent predictor of resilience.Discussion
The first goal of this study was to investigate the existence of group differences in protective factors, resilience, and self-regulation dimensions. Results indicated that girls and students from regular (non-vocational) courses experienced better home environments and had higher resilience levels and impulse control. These results are consistent with previous research, which reported lower self-regulation skills for boys and students who attended vocational courses ( Buckner et al., 2009; GarcĂa del Castillo & Dias, 2009; Jonker et al., 2010; Koning & Boekaerts, 2005; McCabe & Brooks-Gunn, 2007 ). Nevertheless, the existence of gender-related differences in resilience is not consensual in the literature, with some studies indicating that girls and boys react differently to adversity and others concluding that no differences between adolescent girls and boys exist (for a review, see Coleman & Hagell, 2007).
This study reports a disadvantaged pattern of results in home environment, self-regulation, and resilience for students attending vocational courses. The finding that in our sample older students had fewer impulse control abilities may also be related to the fact that vocational coursesâ€™ students were significantly older than the regular coursesâ€™ students. Another important result is related to the fact that, although students from vocational courses reported a better school environment, i.e., more caring relationships, high expectations, and opportunities for meaningful participation in school, than students from regular courses, their resilience levels were lower. This result can be better understood if we look at the individual contribution of protective factors to resilience in the regression models presented in Table 4 . In fact, home environment has the highest effect on resilience, as school environment shows no predictive power. This can indicate that providing good school environments may not be sufficient to improve the resilience levels of the students if they are immersed in home environments lacking supportive interpersonal relationships.
The second and main goal of this study was to test if the self-regulation dimensionsâ€“goal setting and impulse controlâ€“are mediators in the relationship between the four protective factors and adolescentsâ€™ resilience levels. Moderate correlations were obtained between the protective factors and adolescentsâ€™ resilience. However, the results from linear regression modelling indicated that, overall, home environment was the best resilience predictor, followed by community and peer environment. The fact that home environment, namely the interpersonal relationships that are experienced in the family context, is closely related to resilience in children and adolescents is a common conclusion in the literature (e.g., Vanderbilt-Adriance & Shaw, 2008 ). Family ties and the existence of supportive relationships and modelling within family members are frequently referred and valued by adolescents when investigating how they face adversity ( Smokowski et al., 1999 ). Interpersonal relationships and participation in the group of peers and community are protective factors less studied than the interactions established within the family or home setting, although some previous studies have pointed out the importance of friends influence (e.g., Smokowski et al., 1999 ) and community support relationships and resources (e.g., Davies et al., 2011 ) in individualsâ€™ resilience levels. Our results support these findings, highlighting the need for promoting positive interactions in peer and community environments to help adolescents face adversity. A surprising result was that school environment did not predict adolescentsâ€™ resilience levels, because some studies have suggested that meaningful relationships between students and caring adults, the communication of high expectations for academic performance, and the maximization of the opportunities for meaningful participation of students in the school environment can foster resilience (e.g., Brooks, 2006 ). Our results provide a broader understanding of the role of school environment in resilience, suggesting that promotion of supportive relations and opportunities in schools should be accompanied by the fostering of the same quality of relationships and opportunities for participation in other life contexts of adolescents, taking into ac an ecological perspective of human development.
Another finding from this study was that protective factors were weakly correlated with self-regulation dimensions (goal setting and impulse control abilities). In fact, only home environment had a correlation higher than .20 with both self-regulation dimensions. The importance of family environment to the emergence of self-regulation abilities is extensively mentioned in the literature ( Choe et al., 2013; Colman et al., 2006; Moilanen et al., 2010 ) and our results reinforce the importance of this protective factor not only in the overall adaptation to adversity (i.e., resilience) but also in the development of goal setting and impulse control abilities.
Our final finding was that, although goal setting and impulse control abilities are significant predictors of adolescentsâ€™ resilience, these self-regulation dimensions did not mediate the relationship between protective factors and resilience. This means that home, community, and peer environments exert a direct effect on resilience and not an indirect effect mediated by self-regulation. In this sense, home, community, and peer environment should be considered external factors and self-regulation should be understood as an internal factor that contributes to the development of high levels of resilience in adolescents.Limitations and Implications for Future Research and Practice
Our findings should be considered in the scope of some limitations that are present in this study. Firstly, the sampling procedure and the number of missing values limit the generalisability of the results. Future studies should use a larger and stratified sample. Secondly, there is much discussion about the adequate measures to evaluate self-regulation in adolescence. Much of the research in this domain translates models from adults to adolescents, generating theoretical and methodological discussion about the adequacy of the transposition of the measures used to assess adults to adolescentsâ€™ assessment. Despite these limitations, this might be one of the first studies exploring the mediating role of self-regulation in the relationship between protective factors and resilience.
These data are also particularly interesting for intervention, once they highlight the importance of an ecological intervention to promote resilience. As part of the individual intervention, resilience is a multidimensional process that should be promoted with the involvement of parents, peers, and community.Conflict of Interest
The authors of this article declare no conflict of interest.
Copyright © 2018. Colegio Oficial de Psicólogos de Madrid