Professional Documents
Culture Documents
Vol. 4 No. 3 March 2017
Vol. 4 No. 3 March 2017
3; March 2017
Strengths Needs and Resilience among Foster Care Youth: A Measurement Model
Abstract
_________________________________________________________________________
Recent research studies indicate that strengths possessed by underprivileged youth may mediate
the impact potentiated by adverse life circumstances on deleterious developmental outcomes in
Foster Care (Smokowski, Reynolds, & Bezruczko, 2000). As such, this study proposed to explore
the processes by which strengths may act as mechanisms of psychiatric improvement by
delineating the influences of strengths upon developmental outcomes (Gillham, Reivich, & Shatte,
2002). This paper also identifies the main predicting factors of strength based services for youth
in foster care. Effective treatment in foster care formative years often leads to positive
contributions made by youth who overcome disadvantage and mature into generative adults,
while society must devote human and financial resources to assist and remediate youth who
remain limited by adverse circumstances. This paper determines whether pairs of correlations
differed significantly at the alpha level of .05, by using a t-test for differences between strength
and needs dependently sampled correlations.
3
©Research Institute for Progression of Knowledge www.ripknet.org
Table 1. Descriptive Statistics
Std.
Mean Deviation N
RESILIENCY_LOG .6464 .16708 97
CHILDRISKBEHAVIOR 4.7423 3.67502 97
CHILDBEHAVIOREMOTIONALNEED 11.0928 4.76551 97
LLIFEDOMAINFUNCTIONING 12.7526 5.43145 97
CAREGIVERSTRENGTHSNEEDS 10.4330 5.37143 97
CHILDSTRNGTHS 22.1959 6.00805 97
CCULTURE .5670 1.33000 97
SSUCIDERISKMODULE 2.4124 1.28089 97
VVIOLENCEMODULE 3.8660 1.53859 97
EMOTIONALBEHAVIORRISKS 5.9278 2.41630 97
SSABSEXUALLYAGGRESSIVEBEHAVIOR 10.1237 1.10168 97
RRUNWAY 8.0928 2.52109 97
JJUVENALJUSTICE 9.0309 2.07893 97
FFIRESETTING 8.0103 .97355 97
TTRAUMA 7.7938 3.32584 97
SSUBSTANCEUSE 6.3711 1.49527 97
SSCHOOL 5.7010 2.80314 97
DDEVEDLOPMENTALNEEDS 3.1856 1.16664 97
FFAMILYCARETAKER 7.3505 3.43099 97
PPSYCHIATRICHOSPITALIZATIONHISTORY 4.9072 2.17506 97
The initial model (Table 1) hypothesized that Resilience is predicted by the variables listed in Table 1. However,
this initial model, did not have acceptable model fit statistics. Although, regression correlations between most of
the independent and dependent variables were statistically significant (p < .05), only five independent variables
(EMOTIONALBEHAVIORRISKS, RUNWAY, SEXUALLYAGGRESSIVEBEHAVIOR,
CHILDRISKBEHAVIOR, SCHOOL) satisfied the statistical criteria for entry into the regression model. The
independent variable, SCHOOL had the largest correlation (.73) in relation with the dependent variable,
Resilience. CHILDRISKBEHAVIOR had the second largest correlation (.70). Other variables in the initial model
did not have sufficient statistical indices to be included in the multiple regression analysis (Carver & Nash, 2006).
See Table 2.
Table 2. Variables Entered/Removed
Model Variables Entered Method
1 Stepwise (Criteria: Probability-of-F-to-enter <=
EMOTIONALBEHAVIORRISKS
.050, Probability-of-F-to-remove >= .100).
2 Stepwise (Criteria: Probability-of-F-to-enter <=
RUNWAY
.050, Probability-of-F-to-remove >= .100).
3 Stepwise (Criteria: Probability-of-F-to-enter <=
SABSEXUALLYAGGRESSIVEBEHAVIOR
.050, Probability-of-F-to-remove >= .100).
4 Stepwise (Criteria: Probability-of-F-to-enter <=
CHILDRISKBEHAVIOR
.050, Probability-of-F-to-remove >= .100).
5 Stepwise (Criteria: Probability-of-F-to-enter <=
SCHOOL
.050, Probability-of-F-to-remove >= .100).
The model summary statistics indicating the 'goodness of fit' of the model is projected in Table 3. This table
showed the multiple correlation coefficients R, the R-squared (R2) and the Adjusted R-Squared (R2) version of
this coefficient, which can range from 0 to 1 and indicate the 'goodness of fit' of the model.
4
International Journal of Education and Social Science www.ijessnet.com Vol. 4 No. 3; March 2017
Table 3. Model Summary
Std. Error Change Statistics
R Adjusted of the R Square F Sig. F Durbin-
Model R Square R Square Estimate Change Change df1 df2 Change Watson
1 .627a .393 .386 .13091 .393 61.381 1 95 .000
b
2 .658 .433 .421 .12710 .041 6.775 1 94 .011
3 .687c .472 .455 .12339 .038 6.744 1 93 .011
4 .709d .503 .481 .12038 .031 5.708 1 92 .019
5 .730e .533 .508 .11724 .031 5.981 1 91 .016 2.225
Results revealed that the model containing the five variables, (EMOTIONALBEHAVIORRISKS, RUNWAY,
SEXUALLYAGGRESSIVEBEHAVIOR, CHILDRISKBEHAVIOR, SCHOOL) predicted 53.3 percent of the
respondent’s resilience. Applying Cohen's criteria for effect size, the relationship between Resilience Factors and
the five independent variables was characterized as very strong (Multiple R = .73). The multiple regression square
value was .53 and its adjusted square was .58. The model showed that about 50% of the total variation in the
resilience factors of the respondents to be accounted for by a linear combination of the five independent variables
in the model summary (Rosenthal, 2001).
Additionally, the Durbin-Watson statistics suggest that values of test results should range from 1.5 to 2.5. Since
Durbin-Watson results shown on Table 3 are 2.22 it is safe to conclude there is no issue of multicollinearity. The
absence of multicollinearity suggests that another requirement for multiple regression analysis is satisfied, which
increases validity of the multiple regression results. ANOVA tested the statistical significance of the model as
results of displayed in table 4. indicate that the linear combination of the variables,
EMOTIONALBEHAVIORRISKS, RUNWAY, SEXUALLYAGGRESSIVEBEHAVIOR,
CHILDRISKBEHAVIOR and SCHOOL had a statistically significant relationship with resilience (F (1) = 61.381,
p < 0.01) as hypothesized.
Table 4. ANOVA
Model Sum of Squares df Mean Square F Sig.
1 Regression 1.052 1 1.052 61.381 .000b
Residual 1.628 95 .017
Total 2.680 96
2 Regression 1.161 2 .581 35.944 .000c
Residual 1.519 94 .016
Total 2.680 96
3 Regression 1.264 3 .421 27.675 .000d
Residual 1.416 93 .015
Total 2.680 96
4 Regression 1.347 4 .337 23.234 .000e
Residual 1.333 92 .014
Total 2.680 96
5 Regression 1.429 5 .286 20.790 .000f
Residual 1.251 91 .014
Total 2.680 96
The strength of F-values and the p-values been far from (p< 0.05) indicated correlation exists among the five
independent variables in the model, which together, they explain 53% of the variance of the dependent variable,
Resilience as shown in Table 4.
Table 5. shows the beta weights (sometimes called regression coefficients) and the statistical significance
associated with the beta weights. The regression coefficients table include un-standardized regression weight (β),
standard error of estimate (SEβ), the standardized coefficient, the t-ratio, tolerance values, VIF values and level at
which the t-value is statistically significant. The estimated regression coefficients represent levels of the predicted
changes in the dependent variable by each of the independent variable in the model.
5
©Research Institute for Progression of Knowledge www.ripknet.org
Table 5. Coefficients
Standar
Unstandardiz dized 95.0%
ed Coeffic Confidence Collinearity
Coefficients ients Interval for B Correlations Statistics
Std. Lower Upper Zero- Toler
Model B Error Beta t Sig. Bound Bound order Partial Part ance VIF
1 (Constant) .390 .035 11.014 .000 .319 .460
EMOTIONALBEH 1.00
.043 .006 .627 7.835 .000 .032 .054 .627 .627 .627 1.000
AVIORRISKS 0
2 (Constant) .293 .051 5.800 .000 .193 .393
EMOTIONALBEH
.041 .005 .594 7.555 .000 .030 .052 .627 .615 .587 .975 1.026
AVIORRISKS
RRUNWAY .014 .005 .205 2.603 .011 .003 .024 .299 .259 .202 .975 1.026
3 (Constant) .578 .120 4.808 .000 .339 .817
EMOTIONALBEH
.041 .005 .587 7.683 .000 .030 .051 .627 .623 .579 .974 1.027
AVIORRISKS
RRUNWAY .020 .006 .297 3.529 .001 .009 .031 .299 .344 .266 .800 1.250
SSABSEXUALLYA
GGRESSIVEBEHA -.033 .013 -.216 -2.597 .011 -.058 -.008 -.070 -.260 -.196 .820 1.220
VIOR
4 (Constant) .540 .118 4.560 .000 .305 .775
EMOTIONALBEH
.047 .006 .685 8.051 .000 .036 .059 .627 .643 .592 .747 1.339
AVIORRISKS
RRUNWAY .025 .006 .373 4.232 .000 .013 .036 .299 .404 .311 .698 1.433
SSABSEXUALLYA
GGRESSIVEBEHA -.032 .012 -.213 -2.619 .010 -.057 -.008 -.070 -.263 -.193 .819 1.220
VIOR
CHILDRISKBEHA
-.010 .004 -.221 -2.389 .019 -.018 -.002 .239 -.242 -.176 .634 1.578
VIOR
5 (Constant) .546 .115 4.736 .000 .317 .776
EMOTIONALBEH
.045 .006 .652 7.772 .000 .034 .057 .627 .632 .557 .728 1.373
AVIORRISKS
RRUNWAY .024 .006 .365 4.254 .000 .013 .035 .299 .407 .305 .697 1.435
SSABSEXUALLYA
GGRESSIVEBEHA -.037 .012 -.242 -3.022 .003 -.061 -.013 -.070 -.302 -.216 .801 1.248
VIOR
CHILDRISKBEHA
-.012 .004 -.267 -2.905 .005 -.020 -.004 .239 -.291 -.208 .607 1.649
VIOR
SSCHOOL .012 .005 .193 2.446 .016 .002 .021 .323 .248 .175 .820 1.219
a. Dependent Variable: RESILIENCY_LOG
For the independent variable EMOTIONALBEHAVIORRISKS, results indicated a standardized beta weights of
.652, a standard error of .006, and a T-value equal to .652 that was statistically significant at the p < .05 level of
significance. Regression analysis results further indicated the independent variable, RUNWAY had a
standardized beta weight of .365 a standard error of .006, and a T-value equal to 4.254 that was statistically
significant at the p < .05 level of significance.
For the independent variable SEXUALLYAGGRESSIVEBEHAVIOR, results indicated a standardized structure
coefficient of -.242, a standard error of .012 and T-value equal to -3.022 that was negative and statistically
significant at the p < .05 level of significance. Furthermore, the independent variable CHILDRISKBEHAVIOR,
had a structure coefficient of -.267a standard error of .004, and a T-value equal to -2.905 that was statistically
significant at the p < .05 level of significance. Lastly, the independent variable, SCHOOL results indicated a
standardized beta weights of .193, a standard error of .005, and a T-value equal to 2.446 that was statistically
significant at the p < .05 level of significance.
6
International Journal of Education and Social Science www.ijessnet.com Vol. 4 No. 3; March 2017
The independent variables EMOTIONALBEHAVIORRISKS and RUNWAY had the highest structure coefficient
indicating they were the main predictors of resilience. The other independent variables
(SEXUALLYAGGRESSIVEBEHAVIOR, CHILDRISKBEHAVIOR and SCHOOL) had a lower level of
premiums associated with resilience. At any rate, all the correlations between, the independent variables and the
dependent variable were statistically significant. Therefore, we reject the null hypothesis that the slope associated
with the independent variables in the model is equal to zero (b = 0) and conclude that there is a statistically
significant relationship between the independent variables in the model and resilience as hypothesized.
Implications for Practice
The study findings help legitimate the emerging popularity, among providers and consumers, of strengths-based
interventions for youth. Strengths based interventions are derived from the system of care philosophy, a treatment
model that aims to utilize individual and environmental resources in therapeutic processes. Preliminary research
suggests that utilizing strengths in treatment promotes positive outcomes and that strengths-based interventions
may be more effective than traditional, deficit-based services. Nonetheless, questions remain regarding the
clinical utility of strengths. This paper suggests that answers to these questions can be approached by delineating
the pattern of relationships between strengths and needs among at risk youth population.
The excess of one half-million youth in foster care across the United States represents a group at high risk for
undesirable outcomes. Thus, the findings of this study indicate the importance of embracing research-supported
treatment services for foster care youth by providing timely and developmentally appropriate services of sufficient
intensity and duration to build resilient youth. This requires actively coordinating health, mental health, and
educational programs so that assessments are shared a comprehensive yet individualized treatment plan is
developed for every youth. And, consistent with a risk and resilience approach, it requires services that reduce
identified risk factors and promote protection in the least restrictive environment with the least amount of burden
of the foster family.
To change the outcomes for foster care youth, a combination of both early intervention and research studies with
high research rigor is needed. Despite the use of mental health services by the foster care system, the application
of evidence-based treatment is very low, with most of the focus on treatment of sexual abuse (U.S. Public Health
Service, 2000). Little attention is given to the deleterious effects of neglect, and neglected children in foster care
often receive no mental health treatment. There is an urgent need to close the gap between research and practice
in providing evidence-based services to foster care youth. This include providing adequate and appropriate
education and training to foster parents; recognizing and managing the range of health and mental health problems
related to child maltreatment; and creating innovative, more effective prevention and treatment services.
7
©Research Institute for Progression of Knowledge www.ripknet.org
References
Compas, B. E., Hinden, B. R., & Gerhardt, C. A. (1995). Adolescent development: Pathways and processes of
risk and resilience. Annual Reviews Psychology, 46, 265-293.
DeAngelis, T. (2004). Taking action for children’s mental health. Monitor on Psychology, 35, 38-41.
De Carvalho, J. & Schumacker (2012). Developmental Assets Model of High Risk Behaviors, Thriving Behaviors
and Juvenile Delinquency. American Journal of Contemporary Research 2, 1-12.
Dilley, J. B., Weiner, D., Martinovich, Z., & Lyons, J.S. (2004). The validity of the Child and Adolescent Needs
and Strengths (CANS). Manuscript submitted for publication.
Dumont, M., & Provost, M. A. (1999). Resilience in adolescents: Protective role of social support, coping
strategies, self-esteem, and social activities on experiences of stress and depression. Journal of Youth &
Adolescence, 28, 343-363
Gillham, J. E., Reivich, K., & Shatte, A. (2002). Positive youth development, prevention, and positive
psychology: Commentary on Positive youth development in the United States. Prevention & Treatment,
5, NP.
Early, T. J. (2001). Measures for practice with families from a strengths perspective. Families in Society, 82,
225-232.
Hodges, K. (1997) CAFAS manual for training coordinators, clinical administrators, and data managers. Ann
Arbor, MI: Kay Hodges.
Kirkpatrick, L., & Feeney, B. (2007). A simple guide to SPSS for windows. Belmont CA: Wadsworth/Thomson
Learning
Lyons, J. S. (2004). Redressing the emperor: Improving our children’s public mental health system. Westport,
CT: Praeger Publishers.
Mason, W. A., & Windle, M. (2002). A longitudinal study of the effects of religiosity on adolescent alcohol use
and alcohol-related problems. Journal of Adolescent Research, 17, 346-363.
Olejnik, S. and Algina, J. (2000) 'Measures of Effect Size for Comparative Studies: Applications, Interpretations
and Limitations.' Contemporary Educational Psychology, 25, 241-286.
Pedhazur, E. & Pedhazur, L. (1991). Measurement, Design and Analysis. Lawrence Eblaum Associaltes,
Publishers: Hillsdale, New Jersey.
Rosenthal, J. (2010). Statistics and Data Interpretation. Belmont, CA.: Woodsworth/Thomson Learning.
Rubin, A. & Babbie, E. (2010). Research methods for social work. Belmont, CA: Wadsworth/Thomson Press
Schumacker, R. Lomax, R. (2004). A Beginner's Guide to Structural Equation Modeling 2nd Ed. Mahwah, NJ:
Lawrence Erlbaum
Smokowski, P. R., Reynolds, A. J., & Bezruczko, N. (2000). Resilience and protective factors in adolescence:
An autobiographical perspective from disadvantaged youth. Journal of School Psychology, 37, 425-448.
U.S. department of Health and Human Services (2003). The AFCARS report: Preliminary Fiscal Year 2014
estimates as of March 2016. Retrieved October 18, 2016, from www.acf.hhs.gov.