Comparison of Social Support Measurement in Youth Residing in the Community Versus Youth Residing in Foster Care

• In Event: Poster Session 09
  In Poster Session: PS 09 Topic - Social Relationships

Abstract

Introduction: Nearly 687,000 children were represented in foster care in 2018 (DHHS, 2019), and these children have increased risk of mental health and behavioral problems compared to their non-foster peers. Perceived social support (PSS) has been identified as an important protective factor for youth mental health outcomes (Taussig, 2002; Jackson, & Warren, 2000). Children in foster care, however, are removed from their primary sources of social support (biological caregivers) and placed with new caregivers unknown to them, making it likely that their PSS is varied and inconsistent with youth in the general population. Lack of contact with biological parents may also increase reliance on other domains of PSS (e.g., mentors/peers) for youth in foster care. The Social Support Scale for Children (SSSC; Harter, 1985) assesses youth PSS across four domains: parent, teacher, classmate and close friend. While this measure has adequate reliability and validity in community samples, the measurement properties have not been fully evaluated in youth in foster care. The present study provides evaluation of the factor structure of the SSSC in a comparison of youth in foster care and community-dwelling youth.
Methods: The study sample consisted of 214 youth in foster care and 163 youth in Midwestern communities (aged 8-12 years old). Participants completed the SSSC, a 24-item questionnaire to assess the child’s PSS as a part of a larger study on the mechanisms of resilience in children. A confirmatory factor analysis (CFA) and invariance testing was conducted to measure and compare the factor structure of the SSSC in the foster and community samples.
Results: Examination of item distributions revealed a U-shape distribution within the foster care sample as compared to a megaphone response pattern within the community sample. These differences were primarily due to a larger proportion of youth in foster care reporting the lowest category (e.g., “not at all true for me”) in response to questions on PSS (see Figure 1) when compared to community sample participants. In CFA analyses, the model within the community sample demonstrated good fit (χ2(203)= 257.275, RMSEA(.023 - .055)= .041, TLI= .959, CFI= .964), while the model within the foster care sample was less than ideal (χ2(203)= 385.023, RMSEA(.055 - .075)= .065, TLI= .809, CFI= .832). Further, measurement invariance was not obtained in a comparison of the two samples. Based on review of parameter estimates, a modified three-factor model (Peer; Teacher; Parent) within the foster care sample revealed excellent fit to the data (χ2(131)= 188.953, RMSEA(.030 - .059)= .045, TLI= .928, CFI= .939; Figure 2).
Conclusion: The current study identified differences in the factor structure of the SSSC across samples of youth from the community and youth in foster care. Results suggest that youth in foster care may experience social support differently when compared to youth in the general population. For example, distinctions in peer relationships may be less meaningful for youth in foster care when compared to youth in the general population. Finally, measures developed for youth broadly may require modification when used with youth in foster care.

Authors

• Farwah Zaidi, University of Florida
  Presenting Author

• Joy Gabrielli, University of Florida
  Non-Presenting Author

• Amanda Bennett, University of Florida
  Non-Presenting Author

• (Lena) Hsin-Yao Chiu, Pennsylvania State University
  Non-Presenting Author

• Yo Jackson, Pennsylvania State University, University Park
  Non-Presenting Author