Research in foster care: the methodological issues and how the issues are addressed.
Many foster care scholars (Butler, 2000; Colton & Williams, 1997; Colton & Williams, 2004; Kendrick, 1995; Kools, 1997; Pithouse & Parry, 1997; Schofield, Beek, Sargent, & Thoburn 2000; Triseliotis, Sellick, & Short, 1995) define foster care as a substitute placement provided in family setting which is flexible and for temporary basis. The purpose of the care can be modified according to the best interest of the child. At the end of the foster placement, it can be transferred to adoption, family reunification, permanence foster family or residential child care. In short, foster care can be considered as a tailor-made substitute placement for children who are separated from their families. In general, many foster care studies focus on the children especially on their behavioral and emotional problems (e.g. Hegar & Rosenthal, 2009; Linares, Li, Shrout, Brody, & Pettit, 2007; McAuley, 1996). Some are focus on the foster parents especially on their characteristic (e.g. Bates, Baird, Johnson, Lee, Luster & Rehagen, 2005) and strategies or parenting skills (e.g. Lipscombe, Farmer & Moyers, 2003; Pithouse, Hill-Tout & Lowe (2002). Other studies are more interested with social workers on their best practices in arranging and supporting foster care (e.g. Holland, Faulkner, & Perez-del-Aguila, 2005). However, it is claimed that many foster care studies understudied the children's experiences in education (Chapman, Wall, & Barth, 2004) and many male caregivers were left out from the studies (Orme & Buehler, 2001).
Foster care is recognized as a difficult research area because getting suitable participants especially children and youth are not easy. Response rate is expected to be low due to several reasons such as subject missing, lack of cooperation from social workers/agencies, subject declined to participate, the placement is too fragile and not suitable for research, lack of follow up by social workers (Gilbertson & Barber, 2002). Dealing with the bureaucracy of the authorities is certainly the biggest challenge in conducting foster care research. If the related agencies refuse to entertain researchers, no matter how significant the problem statements are and how excellent the research design is, it will be totally useless.
MATERIALS AND METHOD
A brief review was conducted on 24 selected foster care literature. The literature is selected based on its diversity of research design in studying foster care. Method section of the literature and authors' justification on their method selection are the main focus of the review. The review may reflect some characteristics of the content analysis in identifying the common methodological issues. This is because content analysis is commonly used to analysis secondary data such as foster care literature (SabithaMarican, 2005). However this review paper did not thoroughly explore the manifest content and the latent content of the literature. As a result, discussion presented in the following sections is mostly academic discourse. Overall, seven methodological issues were identified. The issues are cross-sectional design, comparison group, adult centric, sample size, standardized instrument, male caregivers, and the dilemma of quantitative vs. qualitative. In addition to that, how these issues were addressed by the respective researchers is presented.
RESULTS AND DISCUSSION
A review on 34 research studies on foster families found that many foster care studies used "single point-intime" or cross-sectional design (Orme & Buehler, 2001, p. 5). Conversely, it is argued that cross-sectional design is not useful in examining causality between variables or difficult to determine "causal order" (Chapman, Wall, & Barth, 2004; Orme & Buehler, 2001, p. 5). To address this issues, it is encouraged to conduct foster care studies though longitudinal design (Orme & Buehler, 2001). For instance, Lipscombe, Farmer, & Moyers (2003) interviewed foster parents at two different periods of time which are two months and one year after the placement. This design enables Lipscombe and colleagues to examine the effect of the parenting strategies on the placement. Another significant example is McAuley's (1996) study on long-term foster care. He gathered data from children, caregivers and teachers periodically for two years. He then found changes of the children's behavior and the children's views toward the placement.
However, longitudinal design is difficult to be implemented because it requires larger funding and consume more time. Indeed, as commented by Andersson (2005), it is difficult for researchers to understand the development occur over a period of time as children's view will change with age. This means even a study shows that a child's behavior has improved in the post treatment period, it is still difficult for researcher to conclude that the improvement is caused by the treatment. There are so many uncontrolled variables may contribute to the change of the child's behavior. The findings gathered from the longitudinal design may not necessary be more valid as compared to cross-section design. The decisive factor whether to select the cross-sectional or longitudinal design is still very much depend on the research's questions and nature of the research environment.
As commented by Orme and Buehler (2001), many foster care studies do not compare between foster children with other children in various aspects such as behavioral problems. Without comparison, it is difficult for the researchers to justify their evaluations when they are measuring the children or the general foster placement. As a result, the comparison group design is gradually becoming popular in foster care studies. For instance, Pithouse, Hill-Tout and Lowe (2002) studied 103 foster children with challenging behavior by dividing them into intervention group (54 children) and comparison group (49 children). Measurements were taken before and after the intervention to target the challenging behavior. However, they found that their intervention was hardly effective. In another example, Zetlin, Weinberg and Shea (2005) studied 120 foster care youths on their mathematic and reading skills by dividing them equally into treatment group (60 children) and control group (60 children). Foster youths in treatment group showed significant improvement as compared to those in control group. These comparisons studies were mostly conducted together with longitudinal design. It enables a better understanding on causality and changes over the time (Orme & Buehler, 2001).
Comparison group is important but it could not be perceived as a compulsory element in a foster care research. Some studies such as case study although do not has comparison group, able to provide in-depth and rich unique findings. For instance, a case study report can presents unique contextual analysis on a 14 year-old boy's experiences in foster care (Szabo&Ritchken, 2002). The findings gathered from the case study can be used to help the boy and indirectly can be used as references by other social workers.
Comparison group can be possibly put aside if foster care research is not intended to measure or quantify the subjects. However, for most quantitative studies, comparison group plays a significant role in ensuring the findings are more convincing and rigorous.
Adult centric studies on children are quite common in foster care research. Studying foster children may not necessary require direct contact with the children. Indeed, social workers can be taken as research participants to seek information of the children (e.g. Holland, Faulkner, & Perez-del Aguila, 2005). Conducting a study on adults is very much easier than a study on children. In New Zealand for instance, it is a complicated process to get children participation in research. This is because the ethic committee may not be aware and sensitive with the children's participation rights and indeed participation rights is not recognized as part of children protection (Powell & Smith, 2009). Getting children to participate in a research is a difficult task because children are viewed as vulnerable individuals (Powell & Smith, 2009). However, participating in research to express their views is actually in line with the spirit of article 12 in the UN Convention on the Rights of the Child (Powell & Smith, 2009). Denying children's rights to participate in research has little substance in protecting them from further harm, indeed it is denying them from accessing better resources for their future development (Powell & Smith, 2009).
More and more foster studies are beginning to seeking opinions from the children. Still, children have very little opportunity to express their feelings about their placements and the broader care system (McAuley, 1996). Under the strong adult's influence in foster care studies, direct feedback from the children in foster care are always not sufficient and not nationally represented (Chapman, Wall, & Barth, 2004). Whiting and Lee (2003) claimed that there are very limited studies on younger children (aged 7 to 12) in foster care. Indeed, many researchers do not include foster care youths as their main subjects of study (Kaplan, Skolnik, & Turnbull, 2009). Furthermore, many studies focus on children were getting data from the adults. This can be seen in Hegar and Rosenthal's study (2009). Although they examined 1415 foster children, some of their core instruments (e.g. Child Behavior Checklist and Teacher's Report Form) were completed by care givers. This implies that some of their findings on children are actually the perception of the adults and do not represent the views of the children.
To address this issue, it is recommended that researchers could conduct more child centric study in foster care. It may not be limited to the children's views towards their placement (e.g. Whiting & Lee, 2003), study on the children's experiences in education is also highly needed (Chapman, Wall, & Barth, 2004; Zetlin, Weinberg, & Shea, 2005). Hence, communication and building relationship skills are important (Powell & Smith, 2009) to seeking consents from the significant adults and trust from the children. The adult-centric studies are still relevant and important, but relatively researchers need to pay more efforts in helping the children to express their concerns to the public and the authorities.
Interview is one of the popular techniques used in child centric studies (e.g. McAuley, 1996; Whiting & Lee, 2003). Interview can be conducted to children as young as 8 years old (McAuley, 1996). However, one of the problems in interviewing children is that children may feel uncomfortable to tell negative views on their placement (Chapman, Wall, & Barth, 2004). They may provide standard answers that they think the interviewer would like to listen. Children are afraid of saying wrong things that might upset the interviewer. This happen because children afraid the interviewer has power to remove them to another foster family or children's institution. Indeed, older children can become very defensive during the interview (McAuley, 1996). An inexperienced interviewer may not able to get any valuable information from the older children because they may just remain silent during the session. Thus, well-trained interviewers are needed.
Chapman, Wall and Barth (2004) commented that large scale of foster care studies is limited. The sample size and sampling method used by many foster care studies are questionable (e.g. McAuley, 1996). The sample mostly refers to foster children as they are the primary clients in the foster care system. A short review on a few foster care studies shows that the sample sizes are ranged from the smallest 19 children to the largest 123 children. Table 1 shows a brief summary of few foster care studies and its sample sizes.
These studies (except Whiting & Lee, 2003) which used small sample size generally did not explain how the sampling was made. The selections of sampling methods are mostly purposive or convenient. Random or probability sampling were not mentioned on these studies. Whiting and Lee (2003) did not purposely select 23 children for interview. The numbers of children for interview were stopped at 23 after the authors felt there was repetitiveness on information given by the children. The repetitiveness of the data gathered determines the number of respondents needed. To address this issue, Orme and Buehler (2001) suggested foster care researchers may consider the probability sampling method in ensuring the generalization of the findings. Indeed, large samples are needed especially for research that using statistical analysis (Orme & Buehler, 2001). In line with that, Chapman, Wall and Barth (2004) studied 320 children, while Hegar and Rosenthal (2009) studied 1415 children. Both studies derived data from the National Survey of Child and Adolescent Well-Being in the U.S. which have fulfilled the requirement of random or probability principle.
Larger sample size is also supported by the McAuley (1996), although he only used a very small sample in his study. Since many agree with larger sample size in foster care studies, then how large is considered enough would be an interesting question. For some researchers from developing countries like Malaysia, they do not have resources like Chapman, Wall and Barth (2004) or Hegar and Rosenthal (2009) to retrieve huge sample from a national survey. If a study is not quantitative oriented and generalization of findings is not important, then sample size could be determined based on the quality of repetitiveness as what have been done by Whiting and Lee (2003). For quantitative oriented foster care studies with clear amount of population, the sample size can fulfill the generalization requirement if it is randomly selected with a require number as suggested in the Krejcie and Morgan's (1970 in Sekaran, 2003, p. 294) table of reference.
According to Orme and Buehler (2001), many foster care studies used their own created measures and do not provide information about the reliability and validity of the measures. This discourages meta-analysis to be conducted as every single foster care study is different and unique. Consequently, this will lead to many different or contradictory interpretations and findings among foster care research. This situation is not supportive for researchers in providing valid advice to the frontline social workers or in convincing the policy makers to advocate desire social change. To address this issue, foster care studies have been gradually using standardized psychometrical measures (Orme & Buehler, 2001). Thus it is getting common to see standardized instruments in many foster care studies (Table 2).
Table 2 shows that Child Behavior Checklist, Youth Self Report and Teacher's Report From are the most frequently used instruments in foster care. From the seven instruments presented, three of them are adult-centric measures which completed by adult. The rest are instruments that seek direct responses from the children. All the instruments are usable for general children in the society for various purposes (e.g. behavior, education, mental health), but none of the instruments is specially designed for children in foster care.
The standardized instruments may not be applicable in qualitative study. Open-ended questions and semi-structured interview are mostly used to seek in-depth information from the participants in foster care studies. Even, observation technique could be conducted to capture the "expressive movement" of the children (Webb, Campbell, Schwartz, &Sechrest, 2006). Thus, standardized instruments are welcomed in foster care research, but it should not become dominative method that overrides other collection methods.
Male Caregivers Were Left Out:
Many foster care studies have unintentionally excluded foster fathers by solely focusing on foster mothers (Orme & Buehler, 2001). This commonly happen when foster mothers are mostly the persons who fill up the CBCL and other standardized instruments in assessing their foster children (e.g. McAuley, 1996). Besides, many foster studies do not present full demographic characteristics of foster parents which are useful for further comparison or meta-analysis (Orme & Buehler, 2001).
To address this issue, special foster care study could be conducted to look into the foster father's perspective. Researchers may approach and encourage foster fathers to fill up the standardized instruments in evaluating their foster children. However this could be a difficult task as studies from residential care found that normally the male caregivers are less motivated to participate in event organized by the welfare agencies (Grupper&Mero-Jaffe, 2008).
Quantitative vs. Qualitative:
Whiting and Lee (2003) stated that more quantitative method was used in foster care studies as compared to qualitative method. Some of the reasons could be that quantitative method is more representative as it analyzing larger sample with standardized instrument. Impressive inferential statistic may looks more convincing as compared to long piece of subjective narration. However, quantitative method does have its limitation. For example, Chapman, Wall and Barth's (2004) quantitative study found that 59% of foster children felt their new schools after placement were better than their previous schools. Chapman and his colleagues were not convinced with the meaning of "better" and suggested in-depth study on this matter. This means they need to have the Qualitative explanation on the word "better" to further understand the children's experiences in school. Open-ended question will be very much helpful in this situation (McAuley, 1996). Qualitative method which is popularly associated with interview technique provides more in-depth and contextual analysis on the data. Qualitative method is very useful in child-centric study in foster care. According to Richards and Morse (2007, p.29), it is helpful in learning "from the participants in a setting or a process the way they experience it..." Besides the interview technique (e.g. McAuley, 1996; Whiting & Lee, 2003, Singer, Doornenbal, &Okma, 2004), other interesting techniques such as observation, content analysis on secondary data are also important. For instance, Curran (2008) conducted a historical investigation on foster children in mid-century Philadelphia. He did not conduct interview or use any standardised instruments, but he just analysed 60 case records from the welfare authorities and reviewed scholarly articles published around that period (1946-1963). Based on these materials, he wrote about foster children's experiences and behaviours for that period.
Quantitative and Qualitative methods have always be a dilemma for many social science researchers. The current foster care research shows that both methods are needed and can be concurrently applied in a single study. This is also known as mixed method. For instance, McAuley (1996) interviewed children and mothers. In the meantime, he also used Teacher's Report Form and Child Behavior Checklist. Another example is Singer, Doornenbal and Okma's (2004) study which used The Child Behavior Checklist and conducted semi-structure interview with children. Mix method represents the strengths of both quantitative and qualitative methods. Findings gathered from mix method will be more logically convincing. Nevertheless, researchers who adopt the mixed method would need to clarify the ontological and epistemological stances in forming the research framework. By conveniently adding in a few interviews into a survey design does not warrant a substantive mixed method study. Indeed, a thorough justification in the philosophical stances is highly required for conducting such design.
Literature on research in foster care have highlighted methodological issues namely the cross-sectional design, comparison group, adult-centric design, sample size and sampling, standardized instrument, male caregivers are left out, and quantitative vs. qualitative. These issues frequently occur in discussion of any foster care studies. As recommended by various foster care studies, future foster care studies may embrace longitudinal design, random sampling, larger sample, more comparison groups and standardized instruments. However, it would not be practical to do so when considering budget, time frame, and most importantly the original objective of the study.
Conducting foster care studies is about exploring the unknown to improving the well being of the foster children; the unknown can be better comprehended through the genuine and substantive research questions. Bigger or larger sample is meaningless if it does not in line with the objectives of the study. It is important to choose the most suitable and workable research method that can verify the problem statements and hypotheses. Prior to the research questions, it is therefore unwise to predetermine the research methods. Following this argument, it is no right or wrong to have cross-sectional design or small sample size and others methods mentioned above in foster care studies. In conclusion, there is still no definite way of conducting an empirical foster care study.
Received 14 November 2013
Received in revised form 24 December 2013
Accepted 28December 2013
Available online 18 January 2014
Andersson, G., 2005. Family relations, adjustment and well-being in a longitudinal study of children in care. Child and Family Social Work, 10: 43-56.
Bates, L., D. Baird, D.J. Johnson, R.E. Lee, T. Luster, C. Rehagen, 2005. Sudanese refugee youth in foster care: The "Lost Boys" in America. Child Welfare, 84(5): 631-648.
Butler, I., 2000. Foster care. In M. Davies (Ed.).The Blackwell encyclopaedia of social work (pp.138-140). Oxford, England: Blackwell Published Ltd.
Chapman, M.V., A. Wall and R.P. Barth, 2004. Children's voices: The perceptions of children in foster care. American Journal of Orthopsychiatry, 74(3): 293-304.
Colton, M., M. Williams, 1997. The nature of foster care international trends Adoption and Fostering, 21(1): 44-49.
Colton, M., M. Williams, 2004. Global trends in foster care. In N. Frost (Ed), Child welfare: Major themes in health and social welfare. Volume III: Child placement and children away from home. (pp: 147-159). London, England: Routledge.
Curran, L., 2008. Longing to "belong": Foster children in mid-century Philadelphia (1946-1963).Journal of Social Theory, 42(2): 425-445.
Gilbertson, R., J.G. Barber, 2002. Obstacles to involving children and young people in foster care research. Child and Family Social Work, 7: 253-258.
Grupper, E., I. Mero-Jaffe, 2008. Residential staff's changing attitudes toward parents of children in their care: Rationale and healing effects on children, parents, and staff. Child and Youth Care Forum, 37(1): 43-56.
Hegar, R.L., J.A. Rosenthal, 2009. Kinship care and sibling placement: Child behavior, family relationships, and school outcomes. Children and Youth Services Review, 31: 670-679.
Holland, S., A. Faulkner, R. Perez-del-Aguila, 2005. Promoting stability and continuity of care for looked after children: A survey and critical review. Child and Family Social Work, 10: 29-41.
Kaplan, S.J., L. Skolnik, A. Turnbull, 2009. Enhancing the empowerment of youth in foster care: Supportive services. Child Welfare, 88(1): 133-161.
Kendrick, A., 1995. Residential care in the integration of child care services.In Social work research findings No. 5. Edinburgh, Scotland: HMSO.
Kools, S.M., 1997. Adolescent identity development in foster care.Family Relations, 46(3): 263-271.
Linares, L.O., M.M. Li, P.E. Shrout, G.H. Brody, G.S. Pettit, 2007. Placement shift, sibling relationship quality, and child outcomes in foster care: A controlled study. Journal of Family Psychology, 21(4): 736-743.
Lipscombe, J., E. Farmer, S. Moyers, 2003. Parenting fostered adolescents: Skills and strategies. Child and Family Social Work, 8: 243-255.
McAuley, C., 1996. Children in long-term foster care. Aldershot, England: Avebury. (Research Report).
McWey, L.M., A.K. Mullis, 2004. Improving the lives of children in foster care: The impact of supervised visitation. Family Relation, 53(3): 293-300.
Orme, J.G., C. Buehler, 2001. Foster family characteristics and behavioral and emotional problems of foster children: A narrative review. Family Relation, 50(1): 3-15.
Pithouse, A., O. Parry, 1997. Fostering in Wales: The all Wales review. Adoption and Fostering, 21(2): 41-49.
Pithouse, A., J. Hill-Tout, K. Lowe, 2002. Training foster carers in challenging behaviour: A case study in disappointment? Child and Family Social Work, 7: 203-214.
Powell, M.A., A.B. Smith, 2009. Children's participation rights in research. Childhood, 16(1), 124-142.
Richards, L., J.M. Morse, 2007. Readme first for a user's guide to qualitative method (2nded.). Thousand Oaks, CA: SAGE Publications.
Sabitha Marican, 2005. Kaedahpenyelidikansainssosial. Petaling Jaya, Malaysia: Prentice Hall.
Schofield, G., M. Beek, K. Sargent, J. Thoburn, 2000. Growing up in foster care. London, England: British Agencies for Adoption and Fostering.
Sekaran, U., 2003. Research methods for business: A skill-building approach (4thed.). New York, NY: John Wiley & Sons.
Singer, E., J. Doornenbal, K. Okma, 2004. Why do children resist or obey their foster parents? The inner logic of children's behavior during discipline. Child Welfare, 83(6): 581-610.
Szabo, C.P., D.A. Ritchken, 2002. Race and family placement: A case report and review. South African Journal of Psychology, 32(4): 60-63.
Triseliotis, J., C. Sellick, R. Short, 1995. Foster care: Theory and practice. London, England: B.T. Batsford Ltd.
Webb, E.J., D.T. Campbell, R.D. Schwartz, L. Sechrest, 2006. Unobstrusive measures (rev ed.). Thousand Oaks, CA: Sage Publications, Inc.
Whiting, J.B., R.E. Lee, 2003. Voice from the system: A qualitative study of foster children's stories. Family Relations, 52: 288-295.
Zetlin, A.G., L.A. Weinberg, N.M. Shea, 2005. Improving educational prospects for youth in foster care: The education liaison model.Intervention in School and Clinic, 41(5): 267-272.
Chan Cheong Chong
School of Social Development, Universiti Utara Malaysia, 06010 Sintok, Kedah, Malaysia.
Chan Cheong Chong, School of Social Development, Universiti Utara Malaysia, 06010 Sintok, Kedah, Malaysia
Table 1: Sample Sizes Used by Foster Care Studies. Sample Author(s) Description Sizes 19 McAuley (1996) 2 years longitudinal study 68 Lipscombe, Farmer and Newly placed young foster youths Moyers (2003) with focus on effective parenting skills 123 McWey and Mullis (2004) Family visitation by birth parents 93 Singer, Doornenbal and 45 foster children and 48 birth Okma (2004) children on behavior problems 23 Whiting and Lee (2003) Views on foster placement 120 Zetlin, Weinberg and Youths on their school performance Shea (2005) Table 2: Standardized Instruments Used to Assess Foster Children. Standardized Instrument Used by Child Behavior Checklist McAuley (1996); Hegar and (parent-report) Rosenthal (2009); Pithouse, Hill-Tout and Lowe (2002); Singer, Doornenbal and Okma (2004) Youth Self Report (child-report) Hegar and Rosenthal (2009) Teacher's Report Form McAuley (1996); Hegar and (teacher-report) Rosenthal (2009) The Rochester Assessment Package Chapman, Wall, and Barth (2004a) for Schools (child-report) The Eyberg Child Behavior Linares, Li, Shrout, Brody, and Inventory (parent-report) Pettit (2007) The Loneliness and Social Linares, Li, Shrout, Brody, and Dissatisfaction Scale Pettit (2007) (child-report) The Child Depression Inventory Linares, Li, Shrout, Brody, and (child-report) Pettit (2007)
|Printer friendly Cite/link Email Feedback|
|Author:||Chong, Chan Cheong|
|Publication:||Advances in Natural and Applied Sciences|
|Date:||Aug 1, 2013|
|Previous Article:||Reflections on researching colonial policy on child welfare services using historical method.|
|Next Article:||Premarital pregnancy and abortion among adolescent.|