Relationships Between Socioeconomic Status Indicators and Proficiency Using NAEP Data

• In Event: Measuring Socioeconomic Status to Contextualize Educational Achievement: A Discussion on Methods, Applications, and Challenges

Abstract

The National Assessment of Educational Progress (NAEP), which is federally mandated to measure SES for the purpose of contextualizing proficiency, has historically used student level National School Lunch Program (NSLP) eligibility as a proxy for SES. However, there are some disadvantages to using NSLP as the main measure of SES. First, SES is generally understood to consist of three components: family income, parental education, and parental occupational status (p. 8, NCES 2012), but NSLP measures only family income. Second, although NSLP is primarily an indicator of income, policies such as the Community Eligibility Provision, which permits eligible schools to provide free lunch to all students, threaten the validity of NSLP as a proxy for income. Due to these and other challenges, the SES expert panel convened by the National Center for Education Statistics (NCES) in 2010-2012 made the recommendation for NAEP to create an SES composite consisting of multiple indicators to represent the “Big 3” components.

One first step towards constructing an index involves understanding the contribution of individual variables that would comprise the index. With a goal of better understanding the associations between contextual variables related to SES from NAEP student and school survey questionnaires and student proficiency, this paper presents mediation analyses to examine the relative importance of SES indicators for different subgroups of interest. In addition to student level SES indicators, researchers have explored school and neighborhood level SES, which have been shown to be associated with student achievement beyond individual student SES (OECD, 2004). The NCES expert panel also recommended using school and neighborhood information in addition to student level variables to create an expanded SES measure. Accordingly, this paper emphasizes differences between effects of student level versus school or neighborhood level indicators.
Results indicate that an index comprised of student NSLP eligibility, school level NSLP eligibility, books in the home, parental education, household possessions, and parental employment status explains a large portion of the White-Hispanic achievement gap in NAEP 2017 mathematics scores, some of the White-Black gap, and nearly none of the White-Asian gap.

Additionally, the relative importance of each component of SES differs by subgroup. Lastly, SES relationships with proficiency are largely driven by student-level NSLP, school-level NSLP, and books in the home.

Authors

• Carol Eckerly, Educational Testing Service

• Paul A Jewsbury, Educational Testing Service