Sunday, January 19, 2020
Marquis BR Salong 13, ML 2 (Marriott Marquis Washington DC)
* noted as presenting author
Background and Purpose: Existing literature has consistently pointed out the poorer academic performance of children from low SES backgrounds compared to children from higher SES backgrounds as well as racial disparities in academic achievement. One source of disparate outcomes that has been identified is testing bias within standardized assessment tools. In efforts to test for racial bias within one assessment tool used within educational settings, the current study examines whether the Elementary School Success Profile (ESSP), a multidimensional, online socioenvironmental assessment tool for 3rd through 5th graders, performs equivalently for low-SES African Americans and low-SES Whites in its measurement of two latent constructs: home education environment (caregiver-report) and academic performance (teacher-report). Methods: The current study relied on secondary ordinal data collected from the parents and teachers of 267 low-income Black and White students. Data were originally collected in 2009 and 2010 from parents, teachers, and students across 13 schools in the southeast region of the United States. To test for measurement invariance within the home educational environment and academic achievement constructs of the ESSP across a low-income sample of Black and White students, a multi-group confirmatory factor analysis (CFA) was conducted using Mplus Version 8.1. Additionally, invariance testing was conducted following Bowen and Masa’s four-step approach and the mean scores for each construct were compared across both groups. Results: Despite a small percentage of noninvariance detected among thresholds, the constructs of home educational environment and academic achievement performed the same for low-income Black and White students. When the mean scores were compared, low-income Black and White students were found to experience statistically similar levels of home educational support but low-income White students were found to experience significantly higher scores in academic achievement. Conclusions and Implications: Prior research studies that examined the factor structure of the ESSP were able to confirm that items on the scales were valid measures of the latent variables, however, validity across race within one socioeconomic group was not confirmed. With measurement invariance established and the potential effects of bias within the ESSP now discounted, the true effects of home educational support on low-income Black children’s academic achievement can be examined in future research. Future studies of measurement invariance among other ESSP constructs as well as among other educational assessments and across other student groups are warranted. This study also demonstrates the importance of exploring the differing effects of poverty across racial groups and, in turn, developing school- and home-based interventions that support the academic success of low-income Black students.