Beyond Free Lunch—Alternative Poverty Measures in Educational Research and Program Evaluation
A paper prepared for the annual meeting of the American Educational Research Association, Montreal, Canada, April 11–15, 2005
Most education studies use a simple and convenient measure of poverty: the percentage of children eligible for free/reduced-price lunch. Although this measure provides the proportion of children coming from disadvantaged backgrounds, it does not capture all dimensions of poverty, such as neighborhood effects.
In the context of a large-scale evaluation, we examine alternative, neighborhood-based measures for poverty (the Dissimilarity Index, the Isolation Index, the poverty level of the school neighborhood, and the percentage of single-parent households with children in the school neighborhood). We investigate the relationship between these indices and the free/reduced-price lunch measure, and then explore the alternative poverty measures’ relationship to student achievement by including them as covariates in multilevel regression models. According to our results, school neighborhoods’ poverty levels, percentage of single-parent families, and degree of poverty concentration as measured by the Isolation Index are significantly related to student achievement; however, another measure of concentrated poverty, the Dissimilarity Index, is not. In addition, we found separate neighborhood effects for 2000–2001 student achievement, hence confirming results from earlier studies that connect neighborhood effects to educational outcomes (Datcher, 1982; Dornbusch, Ritter, & Steinberg, 1991; Aaronson, 1998).