Exploring the Measurement of Student Socioeconomic Status Using a Nationally Representative Data Set

• In Event: Roundtable Session 14
  In Roundtable Session: Using Large-Scale Data Systems to Examine Equity Issues in P-16 Schooling

Abstract

This study uses the Early Childhood Longitudinal Survey-Kindergarten (ECLS-K), a nationally representative database, to explore the measurement of Socioeconomic Status (SES) and its relationship with student achievement. Using Latent Class Analysis and Structural Equation Modeling, two alternate measures of SES are created to supplement the existing SES measures available in the dataset. Student classifications into SES categories are compared and student achievement is predicted from SES using each of the measures. Results indicate that students could be differently identified as belonging to an “at-risk” SES group depending on the measure used. Regression coefficients, while similar in magnitude, would be interpreted differently depending on whether continuous or categorical measures are used. Implications for research in practice are briefly discussed.

Authors

• Emily R. Dickinson, Human Resources Research Organization

• Jill L. Adelson, University of Louisville