Examining the Mentorship Needs of Medically Fragile Foster Parents: A Mixed-Method Approach

Background:

Indubitably, being a foster parent of medically fragile foster youth (e.g., medically-complex foster parents) can be challenging. Thus, public and private child welfare agencies must ensure that foster parents are provided the needed supports to ensure successful foster care placements. Though mentorship programs can be one such support, conceptual models that inform the planning, development, and evaluation, specifically from the perspective of foster parents themselves, are few. The purpose of this study was to explicate a conceptual framework describing effective mentorship programs for medically-complex foster parents. Researchers employed Group Concept Mapping (GCM) with a convenience sample of 26 foster parents in one southeastern state to achieve this purpose.

Methods:

GCM is an integrated mixed-method research approach that utilizes non-metric multidimensional scaling (MDS) and hierarchical cluster analyses (HCA) to analyze primary qualitative data. As a result, pictorial representations of the data (e.g., Concept Map) are generated.

At the outset of the project, qualitative statements were collected via focus-group type brainstorming sessions held a various sites throughout the state. Then, participants were invited to a second meeting to sort these statements into piles (based on perceived conceptual relationships) and rate the statements on two variables: Importance and Feasibility.

Once collected, data were analyzed. First, an aggregate similarity matrix was constructed based on participant sorts. This matrix was analyzed via MDS. In summary, MDS analyzes the structures of participant sort data to come up with coordinates that can be illustrated in a two-dimensional space. That is, each of the qualitative statements collected for this study were represented as a coordinate along an x/y axes. Then, the coordinates derived from the MDS analysis were used as input into the HCA, via Ward’s algorithm. HCA combined the coordinates into clusters, that subsequently formed the Concept Map. Based on this map, bivariate pattern match graphs were computed to investigate ratings for importance and feasibility variables. 

Results:

The MDS analysis of the aggregated participant similarity matrix merged after 11 iterations and produced a final stress value of 0.21, indicating a “good fit” between optimal and actual data configurations. HCA yielded a seven-cluster solution, or Concept Map, related to participant perspectives about effective mentorship programs. Clusters included Recruitment, Matching Practices, and Program Research, among others. Mean bridging values for all clusters ranged from .13 to .79, indicating a high level of consensus among participant sort data. The Recruitment cluster was rated the most important and most feasible to effective mentorship programs. Program Research was deemed least feasible, whilst Matching Practices was rated least important. The correlation coefficient for importance and feasibility was .83, indicating agreement among participants.

Implications:

This study is the first known to the authors to employ GCM as a way to delineate a conceptual framework for effective mentorship programs for medically-complex foster parents. This study offers a pragmatic framework that cannot only guide the development and evaluation of mentor initiatives for foster parents, but can also serve as a foundation for future research and evaluative efforts of these programs.