Constructing Socioeconomic Status Measures Using the Trends in International Mathematics and Science Study Data

• In Event: Measuring Student Home Background in Large Cross-National Studies: Conceptual and Methodological Issues

Abstract

The TIMSS data are the largest multi-year, multi-country, nationally representative source of science and mathematics performance in the fourth and eighth grades. However, the current education literature lacks a systematic discussion on constructing SES measures using TIMSS. In this paper we invite such a conversation and inform it with empirical evidence. To do so we propose a novel approach: we use Demographic and Health Survey Data (DHS), which provide a separate, systematically gathered and validated SES measure. We identify four countries in TIMSS 2007 for which DHS data are available for the same year (Indonesia, Jordan, Ukraine) or an adjacent year (Ghana). We use several existing approaches to measure SES using the TIMSS data and propose a few additional approaches. We compare these various measures against the DHS-based SES measure from the same country and time period, which we treat as a benchmark due to the wealth of information that it contains. While this analysis is unable to identify a superior TIMSS-based SES measure, it shows that not all TIMSS-based SES measures are equal; some approaches may be preferable to others. The analysis also indicates that such approaches need not be computationally complex or demanding in terms of country-specific background variables.

Authors

• Amita Chudgar, Michigan State University

• Chad Lee, Michigan State University

• Thomas F. Luschei, Claremont Graduate University

• Loris Fagioli, Claremont Graduate University