Conceptual and Measurement Issues in Family Sense of Coherence: Applying Research on Family Resilience to Transracial and Inracial Adoptive Families

BACKGROUND. Empirical research has established that Family Sense of Coherence (FSOC) (Antonovsky & Sourani, 1988) plays an important role in for the development and healthy functioning of families. FSOC refers to the cognitive orientations that help family members view their families as comprehensible, manageable, and meaningful. For adoptive families, coping with stressors related to pre- and post-adoptive placement is critical in the long-term well-being of adoptive families. Yet, few studies have considered whether the concept of family sense of coherence can be applied to adoptive families. Nor have studies examined the psychometric properties of the most widely used measure of family sense of coherence—the Family Sense of Coherence Scale (Antonovsky & Sourani, 1988)—with respect to adoptive families. Thus, it is not known whether the conceptualization of family sense of coherence and the structure of the FSOC scale hold for adoptive families in general, and more specifically for adoptive families with different characteristics. In response to these gaps in research knowledge and translation of existing knowledge to adoptive families, the present study examines the factorial validity of the FSOC scale with a sample of adoptive families with different characteristics.

METHOD. Data for the study were derived from two large-scale mail surveys on adoptive families. Samples for each of the studies include adoptive families with a wide range of child, parent, and families characteristics. In total, data on 1,662 families were analyzed. First, confirmatory factor analysis (CFA) were conducted to test the factor structure of the FSOC scale. Second, factorial invariance of the FSOC scale was examined across adoptees' developmental stages (as determined by the ages of adopted youth) and between transracial and inracial adoptive families. After factorial equivalence was established, we explored levels of FSOC across different developmental stages and between transracial and inracial adoptive families.

RESULTS. Although the three-factor model provided the best fit to the data (RMSEA=0.05), extremely high correlations between factors questioned whether the three factors are discrete. The one factor model was reasonably acceptable and was chosen for the baseline model for factorial invariance test across adoptees' developmental stages. The factorial invariance was not achieved across adeoptees' developmental stages. The results of partial invariance tests indicated that six items of the FSOC scale were variant between early childhood and middle childhood/adolescence groups. Excluding the six items, factor mean levels of FSOC significantly decreased as adoptees aged. However, there was no difference in FSOC across inracially and transracially adoptive families with adolescent adoptees.

IMPLICATIONS. This study is the most rigorous to date to have examined the psychometric property of the FSOC and the only one that has tested the FSOC with adoptive families with diverse backgrounds. The results support strong applicability of the FSOC to adoptive families and open a new window to understand family resilience in adoptive families in future. However, researchers should acknowledge some items may be interpreted differently by families with adoptees at different developmental stages. Empirical evidence with a psychometrically sound measurement will contribute to the betterment of adoption research and practice.