Recent data have suggested that child welfare caseloads have been increasing in the US and Canada, and available placements have not kept pace with demand (Ferris-Manning and Zandstra 2003). The average length of stay in care has also been on the rise (Barbell and Freundlich 2001). Despite increases in the use of institutional placements, foster care has remained the most often-used placement. Benefits of foster placements over other alternatives have been identified such as better psychological outcomes for children, greater likelihood of family reunification and enhanced life chances for youth (Kluger et al. 2000; National Youth in Care Network 2001; Thomas 1993). However foster parent associations as well as licensing and regulating bodies noted that the retention of foster parents remained a constant challenge (Gibbs 2005). One significant contributor to foster parent attrition was placement breakdown (Fisher et al. 2000; Leschied et al. 2004).

Several studies have examined the challenges of foster parents and factors that contributed to placement disruption. However there were relatively few studies based on the perceptions of foster caregivers themselves. In this paper, a review of literature and overview of the concept mapping method were followed by the procedures employed and results found in a study of foster parent perceptions of needs for a successful placement. Study results were compared to the literature.

There was evidence in the fostering literature that contributors to a successful foster placement were present both within and outside of the foster family. The characteristics of foster children and foster parents as well as the relationships between them have received some attention in the literature. Additionally there were data on relationships with child protection staff as well as other professionals who provided support, which suggested a connection to placement outcome. There has also been research on the material and training needs of foster parents. Social support from others who were not paid by the system had been identified as a contributor to placement success. Finally, studies with foster parents in Canada employing the concept mapping method found that foster parents report a range of needs, challenges and support services.

Qualities of foster parents who were least likely to have a placement breakdown were been described. Those who were motivated by a desire to help out in the community have been found to display stronger bonds to the children in their home (Cole 2005). Their concern for children was apparent (Buehler et al. 2003) in a responsive (Wilson et al. 2003) but balanced parenting style (Chamberlain et al. 2006). Foster parents who had more fostering experience (Chipungu and Bent-Goodley 2004) and had specialized training were more likely to have success (Kirton et al. 2006). Successful foster parents were warm and child-oriented caregivers (Sinclair and Wilson 2003) who had good knowledge of parenting and employed effective problem-solving and child discipline techniques (Linares et al. 2006; Pacifici et al. 2005). They were willing to integrate the child into the family (Leathers 2006), available to their foster children (Schofield and Beek 2005), satisfied in their roles (Wells et al. 2004), saw themselves as healthy (Harden et al. 2004) and were willing to be involved in child placement and future planning decisions (Bussiere 2006; Gerstenzang and Freundlich 2005).

Relationships between foster parents in the home also had an impact on placement outcome. Foster homes with the highest relationship satisfaction (Orme et al. 2004) lowest relationship stress (Lipscombe et al. 2004) and highest ratings of parent-to-parent cooperation were the most successful (Linares et al. 2006b).

Foster children who were the most behaviorally well adjusted were less likely to experience placement breakdown (Macdonald and Turner 2005). Children with physical disabilities or complex health needs were more likely to experience breakdown (Leslie et al. 2005). Previous breakdown was associated with a greater chance of subsequent experience with breakdown (Wulczyn et al. 2003). However child–caretaker relationship was a very important factor in predicting the outcome of a placement. If in the foster home environment the child felt love, belonging, acceptance, importance and support, the likelihood for placement breakdown was reduced (Griffin 2005). If the child felt that she/he had a secure base (Schofield and Beek 2005) and positive interaction with caregiver the outcomes for the placement were more likely to be positive (Sinclair and Wilson 2003).

Positive relationships between foster parents and social workers, agency staff, as well as other professionals in health care and education contributed to greater likelihood of placement success (Chipungu and Bent-Goodley 2004; Evans et al. 2004; Massinga and Pecora 2004; Monck et al. 2004). Feeling valued and respected by social work staff was linked to placement outcome (Kirton et al. 2006). Valued services included professional consultation (Lyons and Rogers 2004), educational assessment (Zetlin et al. 2003), individual and family therapy before and during the placement (Burns et al. 2004; Chamberlain and Smith 2005). Collaborative efforts between agencies (Altshuler 2006) benefited foster families who found it easier to access needed services (Prince and Austin 2005) when they needed it (Leslie et al. 2005) and where it was convenient (Farmer et al. 2005).

Foster placements were more resistant to breakdown when parents had training opportunities, specialized equipment and financial resources. Timing of the training (including both pre-service and in-service opportunities) contributed to positive placement outcomes (Chamberlain and Smith 2005; Craig-Oldsen et al. 2006). Content varied considerably depending on the children in care at the time and foster parent needs (Morton 2004; Pollack 2006). For example, computer equipment and training has been shown to have a very positive effect on foster families (Finn et al. 2004) who used it to contact social workers and obtain information about parenting-related topics (Finn et al. 2005). Family income did have a significant impact on willingness to care for children who were difficult to place due to behavioral challenges (Cox et al. 2003); those with greater incomes also experienced more successes and fewer breakdowns (Orme et al. 2006; Rhodes et al. 2003).

Social support for foster parents was related to placement outcome. This support had come from many different sources such as friends (Farmer et al. 2005) and family members (birth and extended) as well as the foster family’s own family network (Strozier et al. 2004). Formal and informal support networks have been studied and positive effects shown (Peebles-Wilkins 2003). The positive effects of support from local neighbors (Rosenfeld and Richman 2003) and non-specific community resources for local families were documented (Chahine et al. 2005; Massinga and Pecora 2004).

Studies with foster parents in Canada that employed the concept mapping method identified several needs, challenges and support services. Foster parents in a western Canadian province were asked, “What do you need to be a good foster parent?” (Brown and Calder 2000), and “What would make you consider stopping foster parenting (Brown and Calder 1999). The results indicated that foster parents needed good relationships with social workers, cultural sensitivity, good relationships with other family members, financial support and the right kind of personality. Foster parents indicated that they would consider quitting if they did not get the right support from the child welfare system and staff, felt that they were being taken advantage of by the department, saw that the safety of their family was in jeopardy or experienced a significant disruption to their personal or family life. Foster parents in a central Canadian city who cared for children with disabilities were asked, “What services or supports would be helpful to you?” (Brown et al. 2005). Foster parents indicated that understanding by others, financial support, a good relationship with the school, helpful social workers, professionals with experience in disabilities, information about particular disabilities, medical care, culturally-specific services for Aboriginal children, transitional support and respite were helpful. In another study, foster parents in a central Canadian province were asked the question: “What would make you consider ending a foster placement?” (Brown and Bednar 2006). Results indicated that foster parents considered ending a placement if there was danger to their family, the child could not adapt, the child’s behavior was more than they could handle, the child had complex health needs, the foster care agency was unresponsive, their personal situation changed, their own health concerns became a priority or there was a lack of external support in place.

In summary, there were multiple contributors to placement outcome in foster care. The qualities of foster parents and foster children have received considerable attention in the literature. The relationships that foster parents had with one another and the foster children in their home showed a positive association with placement outcome. Additionally, positive connections with professionals both inside and outside of the child protection system were found to contribute to placement success. There were material supports that helped, including computer technology, training, respite and income. Social support from a variety of sources also appeared to be essential to the success of a foster placement. The cultural sensitivity of support received was identified as important.

Method

The concept mapping method has been used to analyze qualitative data. Those who grouped the data together constructed the resulting concepts or themes that represented all contributions. Each participant viewed all of the data and organized it into groups. By not participating in this procedure the influence of the researcher on how the data were grouped was minimized. What was different about concept mapping compared to other approaches to qualitative data analysis was that the participants generated ideas and grouped all of the ideas generated by all participants. Their groupings were statistically analyzed and combined so that the result equally represented the groupings of each participant.

Concept mapping has been employed in several studies in the human services field. The approach has been employed to examine the underlying assumptions and structure of substance abuse treatment programs (Neff et al. 2006). A recent study of factors contributing to the formation of a therapeutic alliance in counselling has been published (Bedi 2006). The approach has also been used to examine sense of community (Herman et al. 2005), challenges faced by individuals with traumatic brain injury (Donnelly et al. 2005a), needs of children with life-limiting health conditions (Donnelly et al. 2005b) as well as priorities for organizational development (Sutherland and Katz 2005).

Components of Concept Mapping

According to Trochim (1989) there were six steps in concept mapping. The first step was the generation of responses to a question asked of a particular group of participants. Second, redundant responses were removed. They were also edited for clarity. Third, the responses were grouped together by participants into themes Fourth, two types of statistical analysis were applied to the groupings. A decision was made about the best number of concepts. Fifth, the major concepts were identified based on the contents of each. Sixth, the map was used for its intended purpose.

Participants

Participants were obtained through the membership list of the Manitoba Foster Family Network (MFFN). The Network’s membership database was jointly held with the provincial government’s Department of Family Services and Housing. The MFFN was a “non-profit charitable organization focused on grassroots community development opportunities to enhance the lives of foster families” (MFFN 2006).

A description of the study was included in a quarterly newsletter sent to all members before any data were collected. The list of telephone numbers for 1664 licensed foster homes was randomized. Participants were contacted by telephone by a research assistant. Data were collected until a point of redundancy where there were no new responses in five consecutive interviews. In this study redundancy was reached after interviewing 63 foster parents. Twenty-three foster homes declined participation. Participant demographics and fostering experience varied (see Tables 1 and 2). No additional demographic information was collected.

Table 1 Foster parent characteristics (n = 63)
Full size table
Table 2 Fostering experience (n = 63)
Full size table

Research Instrument and Questions

When contacted participants were given a description of the study and informed that their responses would be kept confidential and anonymous. They were also informed that they could choose not to answer a question and withdraw their participation at any time. All who were called gave their verbal consent to participate in a telephone interview that included two open-ended questions. They were asked what would make them end a foster placement and what they needed for a successful foster placement. In a different paper the responses to the question about ending a placement were described (Brown and Bednar 2006). In this paper we discuss responses to the following question: “What do you need for a successful foster placement?” Each participant was asked if she/he was interested in grouping the responses together.

Generation of Statements

A total of 221 statements were made in response to the question. Redundant statements were removed, and the others were edited for clarity and essential meaning (Trochim 1989) in the study using an inter-rater agreement process. Responses by males and females were kept separate. However there appeared to be few differences between responses from males and females so they were combined for the purposes of this analysis leaving a total of 79 statements.

Inter-Rater Agreement Process

Three raters including the author, a graduate student and a teacher reviewed the statements separately, removing redundant items and editing the remainder for clarity and essential meaning. The raters compared their ratings and came to agreement on each statement. For example, the statements “expectations that are clear” and “clear expectations” were considered redundant.

Structuring of Statements

At the time of interview participants were asked if they were interested in the sorting task. A total of 35 who expressed interest were later contacted by telephone. Twenty remained interested and were provided with a randomized set of statements made in response to the question. They were instructed to read through all statements and group them together. When they had completed their groupings, they made contact by telephone and provided their responses. Completed sorts from all 20 participants were received.

Representation of Statements

The concept map was produced from the results of two statistical procedures and followed by a decision about the optimal number of clusters based on conceptual and quantitative fit. Multidimensional scaling placed the statements spatially on a map and cluster analysis placed the points into aggregates. The author utilized the Concept System (Trochim 1987) to perform the analyses and construction of the concept maps.

Multidimensional scaling placed each statement was placed on a map called a point map (Trochim 1989). Statements closer together on the map were more likely to be sorted into the same groups more frequently by participants. Statements that were far apart were more likely to be grouped into different piles more frequently by participants. Hierarchical cluster analysis of the multidimensional scaling results was used to group the statements together. The bridging index (a value ranging between 0 and 1) indicated the relationship between a statement and others on the map. A bridging index close to one for a particular statement indicated that it had been grouped together by participants with other statements in different regions of the map. A bridging index close to zero indicated that the statement had been grouped together with others close by on the map.

The researcher made a decision regarding the most appropriate number of concepts. There was no numerical criterion for the selection of the most appropriate number of concepts. The decision was made based on the conceptual fit of the statements within the various cluster solutions generated and the items contributing most to the uniqueness of each cluster based on bridging indices. Cluster solutions from the default of 11 to 6 were examined before concluding that the eight-cluster solution fit the data best.

Cluster Labels

Participants were asked to suggest names for the themes they constructed. About half of the participants provided labels. Following the analysis the researcher, a graduate student and a teacher reviewed the eight-cluster solution and where labels provided by participants fit the contents of the clusters they were used.

Interpretation and Utilization of Maps

The map provided a summary of the concept mapping process.

Results

Figure 1 is the concept map of statements made by participants. The statements that make up the concepts are listed in Table 3.

Fig. 1
figure 1

Concept map of 79 statements

Full size image
Table 3 Concept items and bridging values for concept map
Full size table

Cluster #1—Personality and Skills

Statements in this cluster focused on the personality and skills of foster parents as contributors to a successful placement. Foster parents indicated that it was important to show the foster children “kindness”, “love”, “commitment” and “patience”. They also reported that successful placements required parents who “understand the foster children” so they could develop “trust between me and the children”. They also needed to “be open minded”, show “flexibility”, have “a sense of humor”, possess “wisdom” and “a big heart”. They also needed to “be prepared for challenges” have a “willingness to get out of my comfort zone” as well show “willingness to make adjustments” when needed. The foster parents also reported that they required “self-awareness” in order to trust their “experience” but to also “know my own skills and limits” such as “know how many foster children I can handle”, as well as have “reasonable expectations of the foster children”. They described the need for “communication skills” and “a daily routine” to “be able to handle violent behavior” and have a “stable home”. Finally, they simply needed “space for the children”.

Cluster #2—Information About Child

Statements in this cluster centered on the need for information about the child as well as agency policies and procedures. Foster parents indicated that they needed “information about the child” such as “medical information” so that they would “know problems the foster child has before placement” and could decide if the placement would be “a good match between my home and needs of the foster child”. Gathering information was not seen as any one person’s responsibility, but rather part of a “team effort” to find the best fit between foster parents and foster children. Participants in this study also wanted to “know what I can do with the foster children and what I can’t” so that there were “clear expectations” on the part of the agency and foster parents.

Cluster #3—Relationship with Agency

This cluster contained statements related to the need for participation with the agency as well as role clarity and good relationships with others professionally involved with the foster child(ren). Foster parents in the study indicated that they wanted “respect from professionals” and to participate in “screening to determine best placement” as well as “involvement in case planning” for a child in their care. They wanted “consistency of team members”, “know what we should expect from agency” as well as “support from the agency”. A “worker who comes to my home to observe and help” and a “worker who backs me” were necessary for a successful placement. Participants also needed to “know what agency expects” and “understand my rights as a foster parent”.

Cluster #4—Personalized Services

In this cluster the statements emphasized the nature of services needed by foster parents to promote successful placements. Foster parents indicated that they needed help at the right time from “helpers who are not quick to judge parenting” but who arrive to “help before it is too late to save the placement”. They described the need for “personalized service” which included “specialized services for foster children with special needs”, “crisis intervention services”, or “someone to sound off to”.

Cluster #5—Community Support

The contents of this cluster focused on the types of community support needed by foster parents. The foster parents in this study described the need for “education” through “courses” or an “online resource list”. They needed local “help for foster child with school work” as well as “mental health services”. Foster parents indicated that they needed “reimbursement for approved purchases” that were “paid on time” in order to have enough “money to met basic needs” and buy “personal items for the children”. Income provided opportunities to “access resources” such as “money for recreational services”. Foster parents also indicated that the types of resources needed varied because “there are different realities for families”. They also described the need for “respite” and a good “relationship with foster child’s biological family”.

Cluster #6—Foster Family Networking

The statements centered on the need for good networks between foster parents. Foster parents identified the need to “get together with other foster parents” and “opportunities for foster children to interact with other foster children” such as “gatherings for foster children”. They also identified the need for an “association for foster parents” where they accessed a “library for foster parents” and attended “workshops for foster children”.

Cluster #7—Supportive Family

The statements in this cluster focused on the need for support in the family for a successful placement. Foster parents identified the need for a “supportive extended family” who saw the value in what they were doing as well as an “extended family who help out”. They helped by sharing their “budgeting skills”, advocating to get “timely treatment for child” or an “adequate per diem”. Foster parents also needed “family members who provide respite” and to have “foster children who will be with us for a long time”.

Cluster #8—Look After Self

Statements in this cluster focused on the need to look after one’s self as a contributor to a successful foster placement. Foster parents reported that they needed to “know when I am getting stressed out” and practice good “self-care” strategies such as “stress management” techniques. They also indicated that they needed to feel independent and “know what we need to get for foster children on our own” as well as exercise some “choice about which foster children to take”.

Discussion

In this section the results are compared to the available literature. Although there were many similarities between the literature and study results, some subtle differences are noted.

There is already considerable data on the type of personality characteristics and skills needed to successfully foster, the need for information about the foster child and system, access to relevant services, the broad range of sources of community support as well as self-care strategies.

The literature is largely consistent with the responses of foster parents related to parent characteristics and abilities. The fostering literature includes references to the desire of foster parents to make a difference in their community and help out families (Cole 2005). There are also multiple references associated with the need for foster parents to be oriented toward the children’s needs (Brown and Calder 2000; Sinclair and Wilson 2003), be organized (Linares et al. 2006a) but flexible (Schofield and Beek 2005) and able to adapt to unexpected changes (Wilson et al. 2003).

In the literature there are references to the need for information about the child as well as the need for policy and procedural information in relation to the agency and system. References on the need for information about the child specifically (Craig-Oldsen et al. 2006; Morton 2004) as well as a variety of child-related and developmental issues (Brown et al. 2005; Pollack 2006) delivered both pre- and in-service (Brown and Bednar 2006; Chamberlain and Smith 2005) are apparent.

There are multiple references in the literature on the service needs of foster parents including attention to the type (Brown et al. 2005; Lyons and Rogers 2004), timing (Chamberlain and Smith 2005), duration (Burns et al. 2004) and frequency (Zetlin et al. 2003).

The topic of funding and relationship to foster placement outcomes is discussed in different ways. Studies examine the impact of increased skill/professional fees on foster parent satisfaction and retention (Gibbs 2005). Others look at the means that the foster family had at the time, independent of their funding from an agency, as a predictor of intent to continue and placement stability (Orme et al. 2006; Rhodes et al. 2003). In the present study foster parents link financial means with access to other needed resources such as recreation programs. This is consistent with the available literature (e.g. Brown et al. 2005). As well, the needs for respite and connection to biological families of foster children receive much attention (e.g. Brown and Bednar 2006).

The literature includes references to the needs of foster parents to have supportive families, including immediate and extended, as well as positive relationships with the birth families of their foster children (Brown and Calder 2000; Griffin 2005; Schofield and Beek 2005). The range of ways that families support one another is not explored in detail. However there are studies of ways that a positive relationship between caregivers contributes to the success of a foster placement (Linares et al. 2006b; Lipscombe et al. 2004).

Finally, there are references in the literature for the need to practice self-care as a prerequisite for effective fostering. Time off from fostering has been shown to have positive effects on placement stability (Brown et al. 2005). The need for foster parents to have control over decisions that affect their families, such as whether a foster child would “fit” in their home was also documented (Brown and Bednar 2006; Wulczyn et al. 2003).

There are also some differences between the literature and study results. Positive relationships with the agency and workers are reported as necessary for a successful foster placement. Participants in the present study, like those in previous research, indicate that they want to be valued by agency staff (Kirton et al. 2006). However the participants in this study also identify the need for consistency and clarity in their role and responsibilities to the foster child, agency and system. This need is not as pronounced in the literature.

There are references in the literature on the benefits of networking opportunities for foster parents, foster children and foster families in general (Farmer et al. 2005; Strozier et al. 2004). The need for an association of foster parents is not apparent in the literature as related to placement outcome. However the foster parents in the present study directly link the presence of an association to placement success.

This study is not without limitations. One limitation is the degree to which the sample reflects the characteristics of the population. No population data are available to make comparisons between the foster parents who participated and all foster parents in the jurisdiction. Sample size is also small relative to the population, and therefore may not adequately reflect the diversity of the sample. While the response rate for the interview is good, several families chose not to participate.

Another limitation concerns the possibility of different needs among foster parents with high versus low years of experience. There is a large range among participants on number of years fostering is high. However not enough interviews are conducted to be able to separate responses meaningfully by years of experience.

There are similarities between the perceptions of foster parents in the present study and the recent literature in relation to needs for a successful placement. The literature and the foster parents in the study identify the need for the right kinds of personality and skills to foster, as well as the need for information about the foster child before and during the placement and a good relationship with professionals inside and outside of the child welfare agency and system. Foster parents also identify the need for individualized supports, depending on the level of foster parent experience and foster child need, which are easily accessible. Foster parents also describe their need for opportunities to connect with other foster parents and families as well as having good support within their own immediate and extended families. The need to take care of self is identified and consistent with the literature.

However, there are two differences between the available literature and the results of interviews with foster parents in this study that point to areas of research that may be further developed. One difference concerns the type of help they receive from others. Although the multiple professions involved with children who are in the care of the state are well documented and the resulting jurisdictional complexities as well as service delivery problems are apparent, the type of service to foster parents is less well explored. Many feel like they are clients of the system but do not want to be treated as “clients”. Those parents want to be treated as part of the team of service providers. An examination of differences in foster placement outcome related to role in the system or agency may make a valuable contribution to the literature.

Another difference between the literature and the study results concerns the connection between a foster parent organization and placement success. Participants in the study see the need for a formal organization for foster parents and run by foster parents as necessary for successful placements. Because such associations exist in many US and Canadian jurisdictions, additional research on both the direct and indirect ways that they have an impact on foster placement outcomes may be a valuable addition to the fostering literature.

There appears to be considerable overlap between researcher-defined constructs found within the fostering literature and the perceptions of foster parents on the front lines of service delivery. However there is a difference between what is needed and what is received. There is less literature on the barriers faced by parents in obtaining the necessary external resources and developing the necessary internal family resources to foster successfully.

The degree to which policies reflect the needs that foster parents identify will vary from jurisdiction to jurisdiction. Yet the literature does point to at least two issues for which effective policies may support successful foster placements. One area that has been identified is the need for matching foster families with foster children. Too often in practice the need to place a child overrides the need to find the right kind of foster family for a particular foster child. Another important policy issue concerns information that foster parents can access about the foster children in their care. While there may be privacy reasons for withholding information about a child in care from the foster parents the need for foster family safety (including the safety of the foster child) should be a high priority.