Not All Developmental Assets Can Predict Negative Mental Health Outcomes of Disadvantaged Youth : A Case of Suburban Kuala Lumpur Rozmi

This study sets out to gauge the relationship between developmental assets of adolescents and their negative mental health outcomes. Sample size comprised of 346 respondents of disadvantaged (at-risk) youth from suburban areas of Kuala Lumpur by using purposive and snowball sampling techniques. Two instruments were used; General Health Questionnaire (GHQ-12) by Goldberg and Williams © (1988) to measure negative mental health outcomes (depression/anxiety and social dysfunctions) and a shorter version of Developmental Assets (© Search Institute) to measure developmental assets. Correlation and multiple regression were employed to test the hypotheses through SPSS (V.22.0). Results depict none of the internal assets had significant correlation with negative mental health outcomes, and external assets; particularly, the hope and expectations of family members have significant negative correlation with depression, anxiety and social dysfunction. While, positive peer influence and neighbourhood religiosity had significant negative correlation with overall negative mental health outcomes. On the other hand, hope and expectations of parents toward their adolescents can predict the severity of negative mental health outcomes. Conclusively, findings exhibit positive influence of peers and neighbourhood religiosity are inversely proportional to negative mental health outcomes. The implications of this study would suggest good and timely parental monitoring of the adolescents is a significant contributor in positive youth development.


Introduction 1.
Positive youth development (PYD) initiated as a conceptual domain of developing assets within youth as compared to eradicate risk factors that focus on 'deficit-oriented domains' (Lerner, Brentano, Dowling & Anderson, 2002).Plethora of research indicates an association exists between the number of assets possessed and the number of flourishing indicators within an individual (Enfield & Owens, 2009).Considering this evidence, PYD deliberates the strengths of youth and values that contribute towards healthy development by taking advantage of these individual strengths through evocative societal roles and community-based actions (González et al., 2013).
Community-based activities may serve as protective as well as risk factors that can promote or hinder the positive growth of youth.The youth with increased developmental assets are less likely to exhibit aggressive behaviour and violence (Aspy et al., 2004); smoking (Atkins, Oman, Vesley, Aspy & McLeroy, 2002); unsafe sexual behavior (Oman, Vesley, Aspy, McLeroy, Luby, 2004;Vesly et al., 2004), drug and alcohol use (Oman et al., 2004).
On the other hand, community has direct influence to develop risk factors for youth that contribute in development of negative mental health outcome (O'Connell, Boat, & Warner, 2009).Especially, youth living at disadvantaged locations such as urban poor has a greater risk of negative outcomes as compared to others.According to Moore et al., (2006) youth "at risk" would always been vulnerable to be victimised or be offended by environmental, social and family condition that negatively influence their personal growth and development.There is likelihood, this youth would turn into truants, and early drop outer from school.
In case for Malaysian youth, according to Ministry of Health Malaysia (2011), they are suffering from severe mental health problems, job related stress, and high risk of psychosocial problems.National Health and Morbidity Survey (2014) also indicates 20% youth is facing adversity in terms of poverty, homelessness, environmental risks, negative peer pressure, and unemployment that leads toward negative mental health outcomes.Similarly, Teoh (2010) reported 18.4% male and 16.8% female students are in clutch of clinical aggression and 14% male and 20.6% female are clinically diagnosed as depression patients.
Mental health of youth and developmental assets are linked side by side; developmental assets work as a protective factor against negative mental health outcomes (Benson, et al., 2004).Studies have shown that there is a relationship between the number of thriving indicators within an individual and the number of assets possessed (Benson, et al., 2006).Positive youth development is usually measured through the Search Institute's Developmental Asset Framework (Scales & Leffert, 2004).This framework consists of forty primary assets that may affect healthy youth development.These assets are further divided into two groups; internal or external assets.On the one hand, internal assets based on skills, values, and commitments that stem from within an individual including humility, appropriate decision-making, and a sense for purpose in life.Internal assets comprised of commitment to learning, positive values, social competencies, or positive identity.On the other hand, external assets nourish outside of an individual.They are positive experiences and interactions gained from one's family, non-parental role models, school, community, and service groups.External asset categories include support, empowerment, boundaries and expectations, or constructive use of time (Sesma & Roehlkepartain, 2003).Finally, through developmental assets individual's potential can be maximised to play meaningful and better roles toward positive development of community.
While developmental assets appear to be protective against engagement in risky behaviours, but little is known about the relationship between developmental assets and mental health of the disadvantaged youth.According to Diener (1999), improving quality of life of individual is important for enriching an individual's overall well-being over time including as well as physical and mental health.There are reasons to believe that these assets can affect the health, whether it is school engagement (Bond, et al., 2007), resistance (Parto, 2011), peers (Walsh, et al., 2010), families (Rothon, Goodwin, &Stansfeld, 2012), andneighbourhood (van Voorhees, et al., 2008).In light of these established relationships between developmental assets and youth quality of life, little is known about the relationship between mental health and developmental assets in the context of youth living in suburban community.While studies by Aspy et al., (2004) and Vesely et al., (2004) found significant positive relationship between increased developmental assets and life satisfaction, the findings are limited to one study of public middle school students.Whereas, indigenous studies such as National Health and Morbidity Survey (NHMS, 2014) and Teoh, (2010) did not focus on developmental assets and negative mental health outcomes.Conclusively, this study will bridge the gap between roles of developmental assets and their contribution towards mental health outcomes of disadvantaged youth living in suburban localities of Kuala Lumpur.

Context of the Study 2.
In this study Kuala Lumpur was chosen because it is a complete urban area, where the central zone is largely urban and some rural areas in the northern region.Kuala Lumpur is the main city and capital, located in the Federal Territory of Kuala Lumpur Malaysia, administered by the federal government.About 25% of the population live in over 148 squatter settlements and low-cost housing projects whereby the incidence of antisocial behavior and mental health problems frequently reported involving adolescents from this community.This study was designed to recruit young people identified as 'at risk' by the local community as they may be involved in antisocial behavior.The study was conducted in collaboration with the squatter resettlement of Local Communities, Kuala Lumpur City Hall and the Youth Empowerment Center, Universiti Kebangsaan Malaysia.Respondents were selected from three Community Housing Project, namely, PPR Pantai Ria, PPR Seri Cempaka, and PPR Seri Pantai.Participation in this study is voluntary and written consent was taken from all the participants.Methodology 3.

General Health Questionnaire (GHQ)
General Health Questionnaire was developed by Goldberg and Williams © (1988) to measure the mental health of adults.Initially it consists of 60 items targeting adults and has undergone various adaptations.Most notable version GHQ-12 is being used in this study.Scoring can be done in one of the two ways: Likert score from 0 to 3 or through bimodal response scale (0, 0, 1, 1).This study preferred the first scoring method; the higher the score, the mentally unhealthier the individual is.Respondents are considered unhealthy if they scored at or above 14.The instrument has consistently produced high reliability (r>0.7), and studies (Tait, French, & Hulse, 2003;Baksheev, Robinson, Cosgrave, Baker, & Yung, 2011) proved it as valid instrument to measure negative mental health outcomes.Items 1-6 measured depression/anxiety, 7-12 measured social dysfunction, and the sum of both measured the mental health outcomes of the respondents.The Cronbach Alpha for current study is 0.757.

Developmental Assets (Social Assets Scale)
Developmental Assets (© Search Institute) were initially measured through 40 assets.The social assets in this study are Support, Empowerment, Boundaries and Expectations, Constructive use of Time, Commitment to Learning, Positive Values, Social Competencies, and Positive Identity.The first four of those groups were considered external assets, and the remaining four is the internal assets.External assets measured the external factors contributing to the development of an adolescent, such as the living environment, influence from families and friends, and school involvements.Internal assets are more directed towards the adolescents themselves, measuring factors such as their resistance from negative peer pressure, self-esteem and integrity.Each of these assets is measured in a Likert scale of 1-5 (Strongly Disagree to Strongly Agree).
While the developmental asset framework is comprehensive and reliable, it is apparent that it can be too lengthy and not appropriate for local use.Thus, this study used a modified version developed by local researchers Abdul Kadir et al., (2012).After conducted a pilot study, it was found that 13 more items does not produce acceptable Cronbach Alpha coefficients (less than 0.40).This left the developmental assets with 10 items: school engagement, caring, and resistance for internal social assets; and family support, family communications, hope & expectations of parents, positive peer influence, neighbourhood religiosity, safety of neighbourhood, and care towards neighbourhood for external social assets.School engagement means how the respondents feel comfortable in their school environment.Caring is about how much the respondents care for the people close to them, and resistance is how good the respondents are in resisting the negative influence of their peers.Family support measures the support given by the family or parents.Similarly, family communications is how frequently the family members communicate with each other.Hope and expectations is the parents' expectations toward their child's achievements.Positive peer influence measures the influence by the peers on the respondents.Neighbourhood religiosity aims to see how much the community in the respondents' living environment involves in spiritual activities.Safety measures how safe the respondents' communities are.Finally, the care towards neighbourhood means the frequencies of respondents' involvement in neighbourhood activities, such as cleaning the environment.These assets were measured on a 1-5 Likert scale (1=Strongly Disagree and 5 = Strongly Agree).

Sample
The sample of this study solely based on at-risk (disadvantaged) urban youth living in periphery of Kuala Lumpur in lowcost houses.Researchers seek the assistance from local community representatives to identify the families of at-risk youth.Afterwards, researchers asked the respondents to recommend more individuals fitting the same criteria to reach sample size.An informed consent form was given to the respondent prior to completing the questionnaire.Total 358 questionnaires were collected from prescribed respondents.However, after analysis, we found that only 351 respondents showed reliable data that can be analysed properly.The remaining 7 respondents were rejected because of missing and unreliable data (all items in a scale were answered with the same answer).Initial analysis found unsatisfactory results and did not represent a normal curve.Through box-plot and stem-and-leaf plot analysis, few respondents removed, leaving a total of 346 respondents proceeded for data analysis.

Analyses
After collection of data, demographic results were tabulated along with descriptive analysis.Results from GHQ-12 and Social Assets Scale were also described on a per-scale basis, based on the mean, the median, and the minimum and maximum values.
Correlation analysis was done between independent variables (Social Assets) and dependent variables (GHQ-12, Depression/Anxiety, and Social Dysfunction).Formerly, multiple regression analyses were employed with independent variables that shows significant correlations which were set as the predictors of the dependent variables.All analyses were done through IBM Statistical Package for Social Science, Version 22.0 for Microsoft Windows.

4.
The demographic results were calculated that are shown in Table 1.Majority of Respondents were male (Malay Muslim), aged from 12 to 24 (mean=16) and from a nuclear family.For their living environments, majority lived in low-cost apartments or flats near the Kuala Lumpur city centre.In terms of education, majority of respondents managed to stopover up to middle secondary school.For their household income, the average income is RM1, 231.38, however, this could be skewed since the minimum income was as low as RM150 and the maximum was as high as RM10, 000 and 35% of the respondents did not fill up this part because they do not know their household income.According to Ministry of Women, Family and Community Development (2014) this amount can only bear basic needs such as food and clothing.In terms of respondents' mental health status as measured by GHQ (See Table 2) the score indicating the mean of 14 (SD = 5.99) out of a possible maximum score of 36.Of the 346 respondents, 54% were scored at the lower level, while the remaining 46% scored above median.For the depression/anxiety component, the mean score is 6.13 (SD = 3.61), with the median score of 6.Of the 346 respondents, 58% scored at below, while the remaining 42% scored above median.For the social dysfunction component, the mean score is 7.87 (SD = 3.65), with 57% scored below, while the remaining 43% scored above median.These data indicates that about 40% of respondents in this study scored above median for mental health, either from the total score of GHQ-12 or of its components.Thus, nearly half of the respondents from this study may be considered as mentally unhealthy.
For the Developmental Assets Score, it was divided into internal and external social assets.Three internal social assets analysed which are school engagement, caring, and resistance.School engagement shows a mean of 14.06 (SD = 3.347) with 52.6% scored below while the remaining 47.4% scored above median.For the Caring, the mean score=13.83(SD = 3.254) which 52.9% respondents scored below and the remaining 47.1% scored above median.For the Resistance, the mean score=17.49%(SD = 3.762) with 58.7% respondents scored below and the remaining 41.3% scored above median.These indicated that majority have low score in internal assets (See Table 2).For the external assets, this study measure seven components of external assets which are family support, family communication, hope and expectation, positive peer influence, neighbour religiosity, safety and care toward neighbourhood.For all these components it seems that majority of respondents' score are low (See Table 2).Another interesting finding from this study is the significant correlation was found between the variables measured.The results of correlations analyses between three internal social assets and GHQ was tabulated in Table 3.For the external social assets, only hope and expectations of family members toward the adolescent has the most significant negative correlations for all three dependent variables.Positive peer influence shows significant negative correlations with GHQ-12 and social dysfunction component, while neighbourhood religiosity also has significant negative correlations with GHQ-12 and social dysfunction component.The study also investigated which factors that could predict mental health of young adolescent in this poor community.To answer this multiple regression analysis was performed.Table 4 shows the multiple regression analysis of the total GHQ-12 and its correlates.Result as show in Table 4 indicates that the hope and expectations of parents toward youths, their positive peer influence and the neighbourhood religiosity predict the sum total of general health as much as 2.9% from the total variance.Regression coefficient shows that only the hope toward adolescents ( = -0.135,t (345) = -2.292,p<0.05) predict mental health.It is also found that positive peer influence and neighbourhood religiosity does not predict mental health, thus not support our hypothesis.Based on this finding, the results can simply be understand as below; H=19.416 -0.311hWhere H = the occurrence of mental health according to GHQ; and h = Adolescents' parents' hope and expectations toward their child Further analysis was also performed for the GHQ-12's sub-components especially depression/anxiety and social dysfunction.Although the results of regression analysis did not show any significance in predicting the results.

5.
This paper exmined to what extend developmental assets predict mental health outcomes of disadvantaged adolescent in Kuala Lumpur sub-urban.Evidence of socio-demographic characteristics showed that most respondents were came from low socioeconomic background family.The family income was RM975-per month, is considered the bare minimum for basic living (RM830-per month) as specified by Ministry of Women, Family and Community Development (Ministry of Women, Family and Community Development Malaysia, 2014).About 66% of repondents live in low-cost house or apartments near KL city centre, where the cost of living is higher.These factors may affect the adolescents mental health and wellbeing as it has been proved that low socioeconomic status can reduce their quality of life (Thumboo, et al., 2003) and they are more prone to behavioral problems (Reijneveld, et al., 2010).
From the descriptive analysis, it was found that 46.5% of respondents were above the score (rated by GHQ-mental health status) considered unhealthy by Goldberg and Williams (1988).In contrast, according to the National Health and Morbidity Survey (2012) by the Ministry of Health Malaysia, the rate of mental health issues among teenagers was 20%less than half.Various variables could take into account, particularly the demographics, as this study focused on at-risk youth in suburban Kuala Lumpur and cannot generalized on the overall population.
Meanwhile, the correlation analysis shows only three assets have significant negative correlations with mental health outcomes: hope and expectations of family members toward respondents; positive peer influence, and neighbourhood religiosity.As correlation depicts the relationship between values and does not imply a direct causation, what can be surmised from these results are that the higher the expectation of family members toward respondents; or the more positive the influence from their peers; or the more religious the neighbourhood in their living environment, the less likely the respondents will have negative mental health outcomes -and vice versa.For more accurate results, multiple regression analysis was run on each of the three dependent variables.
Results from multiple regression analysis found that only the hope and expectations of the family members for the achievements and success of their youth can reliably be predicted for positive mental health outcomes of youth.It seems that when the family members, especially parents, believe in the potential of their sons and daughters, and monitor their progress and encourage them, they will strive better.The opposite can also be said: when an adolescent does not have the proper care from their parents, their mental health decreased.When an adolescent argues with their parents and worried about their familial relationships, their GHQ score increases (Sweeting, West, Young, & Der, 2010).Similarly, adolescents who does not experience parental warmth will have lower self-esteem and higher tendency to experience depression (Jun, Baharudin, & Jo-Pei, 2013).Depression may ultimately lead to suicide, and adolescents whom are more attached to their parents were less likely to attempt suicide (Maimon, Browning, & Brooks-Gunn, 2010).
Finally, it is predicted that family connectedness, parental involvement and warmth protect their adolescents from negative mental health outcomes such as depression; this notion is also endorsed by Van Voorhees et al. (2008).Similarly, adolescent who has supportive parents score better academically (Rothon, Goodwin, & Stansfeld, 2012).

Conclusion 6.
This study found that external assets such as low expectations and hope from parents and family members may have impact on adolescent's depression, anxiety and social dysfunction.Thus the implication of this study suggested that parental factors such as parental warmth, autonomy giving, care, and timely supervision can contribute in positive development of youth living in suburban areas of Kuala Lumpur.Although, it is too early to conclude and generalise this finding as this study may have some limitations.For example the sample may not represent the actual population of young adolescent and the area covered just only in Kuala Lumpur suburban.Future researchers should take account on these drawbacks and future studies should focus on parental factors as it is suggested that Malaysian youth is significantly influenced by their parents, so this construct needs serious consideration for better understanding of youth development.

Table 1 :
Demography of respondents