Assessing Family Outcomes Over Time: Longitudinal Measurement Invariance of the Family Outcomes Survey–Revised

Abstract

The Family Outcomes Survey–Revised (FOS-R) measures family outcomes of services for young children with disabilities or developmental delays, and can be used to investigate change in outcome achievement over time. To confirm the validity of FOS-R change scores, it must be established that the FOS-R measures the same construct over time; if so, score changes can be attributed to family outcome achievement. However, if not, interpretation of change scores may be misleading. Caregivers ( N = 349) of Singaporean children with developmental delays receiving early intervention services completed the FOS-R Part A twice, approximately 1 year apart. Factor analysis was used to confirm the structure of the FOS-R Part A and to assess whether the scale measures the same thing across time (i.e., longitudinal measurement invariance). Results provided support for partial longitudinal measurement invariance, indicating that change scores on this scale can be validly interpreted with certain caveats.

Measuring family outcomes for young autistic children receiving interventions in China: The structure of the Family Outcomes Survey and predicting variables

“…Another feature of this five-factor modified model was the correlated errors of items 14 and 15, which is different from what has been reported in previous studies of the FOS-R-A (e.g. Waschl et al, 2021). From the wording used in the items, one may infer why these items might be characterized by greater shared variance than that explained by Outcome 4 (having support systems).…”

“…This provides new evidence for the cross-cultural applicability of the FOS-R-A structure in China. The fact that Model 5 showed a better fit than the second-order models (M2, M4, and M6) is consistent with Waschl et al (2021) and supports the use of the five family outcomes as independent subscales.…”

“…Moreover, Model 5 is slightly different from Model 3, which has been reported to show the best fit with a Singaporean data set (Waschl et al, 2021). This is indicative of the impact of the cultural context on family outcomes.…”

“…A robust maximum likelihood estimator, which is appropriate for non-normal data distributions, was used. Factor structures reported in previous studies (Bailey et al, 2011;Poon et al, 2014;Waschl et al, 2021) were tested: (1) five factors (understanding your child's strengths, needs, and abilities; knowing your rights and advocating for your child; helping your child develop and learn; having support systems; and accessing the community); (2) five factors with a higher order general family outcome factor; (3) five factors with a mini testlet factor to account for the shared variance of items on the fifth factor measuring practical aspects of care; and (4) five factors with a mini testlet factor and a higher order general family outcome factor.…”

“…There is an emerging consensus (Waschl et al, 2021; Wicks et al, 2021) on the need to measure parental and family functioning to better understand the impacts of EI services in a wider context. A family-centered approach to EI services requires addressing family outcomes in various dimensions, such as focusing on family strengths, respecting family diversity and values, encouraging family decision-making and empowerment, communicating with families in an open and collaborative manner, adopting a flexible approach to service provision, and recognizing the value of informal support systems (Bailey et al, 2011).…”