Comparing Social Support and School Belonging in Black/African American and White Students: Measurement Invariance Testing of the Essp-C

Background: Social support and school belonging are considered key constructs in school social work research because of their associations with a host of positive outcomes, including increased self-esteem, better problem-solving skills, greater academic achievement, and higher levels of school engagement. A limitation of incorporating social support and school belonging into school social work interventions is that many widely-used measures of the constructs have not been validated for use with diverse populations. The purpose of the current study is to determine whether an existing measurement tool assessing school belonging and social support (the Elementary School Success Profile for Children; ESSP-C) performs equally for both Black/African American and White children, and can thus be used to make valid assessments and comparisons between groups.

Methods: Responses to 28 ESSP-C items assessing levels of school belonging and social support were collected from 555 elementary school students. 54% (n=301) of the sample identified as White, and the remaining 46% (n=254) identified as Black/African American. Students were diverse in terms of sex, socioeconomic status, and geographic characteristics (urban vs. rural).

To evaluate the suitability of the ESSP-C items for cross-group comparisons, a multiple group confirmatory factor analysis was conducted in Mplus 7.1. Because data collected with the ESSP-C are both non-normal and ordinal, weighted least squares means and variances (WLSMV) estimation was used for the analysis. Invariance testing proceeded according to the steps established for WLSMV estimation (Byrne, 2012; Muthén & Muthén, 1998-2012; Sass, 2011).

Results: The selected ESSP-C items demonstrated configural invariance between White and Black/African American students, indicating that the basic structure of the social support/school belonging model is the same for both groups. Although full measurement invariance did not hold, the limitation of non-invariance to a very small percentage of parameters means that the ESSP-C can be considered to provide valid assessment of social support and school belonging for both Black/African American and White students. The establishment of partial measurement invariance also allowed comparison of latent means. No statistically significant differences existed in levels of family support, friend support, teacher support, and general social support between groups. However, Black/African American students did show a statistically significantly higher level of school belonging compared to White students.

Implications/Conclusions: Establishing that the ESSP-C can be used to make valid comparisons of the levels of school belonging and different forms of social support reported by White and Black/African American children brings a new use and interpretation of ESSP-C data, as well as contributes to general measurement and intervention research incorporating school belonging and social support. In addition to its ability to assess multiple domains of a student’s general social environment and identify potential barriers to learning, the ESSP-C can also be used to highlight potential strengths or challenges within a student’s social network. Awareness of these strengths and challenges can then be used to guide school belonging and social support interventions, including use as a pre- and post-test for intervention evaluation.