New Student-Poverty Measures Proposed for National Tests
Proposed indicators go broader, deeper
Aiming to get a clearer picture of how students' home and community resources affect their academic achievement, America's best-known K-12 education barometer, the National Assessment of Educational Progress, is building a comprehensive new way to gauge socioeconomic status.
The new measure, being developed by the National Assessment Governing Board and the National Center for Education Statistics, is intended to look beyond a traditional measure of family income to a child's family, community, and school supports for learning.
"This issue has just been on the burner for so, so long," said Maria V. Ferguson, the executive director of the Washington-based Center on Education Policy. "When NAGB starts talking about it, that does elevate it to a place where it could be part of a bigger policy debate," she said. "I wonder if the folks at NAGB are hoping this could be an opening salvo into a bigger conversation about how [different SES measures] might affect other programs."
The governing board commissioned eight researchers in education, economics, statistics, human development, and sociology that have been working on the new indicators since 2010. The panel released its initial proposal at a NAGB meeting here Nov. 29.
"We rapidly learned that socioeconomic status contains multiple dimensions and categories that don't neatly collapse back to 'low' versus 'high,' " said Charles D. Cowan, the chief executive officer of the San Antonio-based research group Analytic Focus and a member of the governing board's expert panel. "Over the last 10 to 15 years, there's been an explosion in the data available" on student characteristics, Mr. Cowan said. "Perhaps now is the time to think about alternative measures of SES simply because now we are able to think about it."
Beyond Free Lunch
The National Assessment Governing Board is considering a new method of identifying a student’s socioeconomic status when disaggregating the results of the National Assessment of Educational Progress. NAEP researchers now rely primarily on a student’s eligibility for the National School Lunch Program—which as of 2011 provided free or low-cost meals to more than 31 million students in poverty each day—as a proxy for socioeconomic status. This traditional indicator is bolstered by background questions on home possessions, such as washing machines, encyclopedias, and mobile phones.
Proposed New “Core” SES Indicators
• Family income and indicators of home possessions and that have been shown to be linked to educational access, such as Internet availability and number of books in the home
• Parents’ educational attainment
• Parents’ occupational status
Potential “Expanded” SES Indicators
• Family: For example, family structure, stability, and the presence
of extended family and other supportive adults
• Neighborhood: Including the concentration of poverty or linguistic isolation, the percentage of unemployed adults, and the availability of museums, parks, or safe walking routes
• School: The aggregate SES composition of students at the school the child attends, as distinct from the neighborhood SES level
Potential Additional Context Indicators
• Physical stressors: Local rates of illness or environmental problems
• Psychological stressors: Levels of crime in the school and community
• Psychological protectors: Student perception of parent involvement and expectations
SOURCE: National Assessment Governing Board
For decades, the universal proxy for students' socioeconomic status—for NAEP and nearly every federal education and child-health program—has been just such a high-low indicator: eligibility for subsidized meals under the National School Lunch Program.
Federal food aid does capture a huge swath of students in poverty: The school lunch program alone provides meals for more than 31 million children, at reduced cost to those living at or below 185 percent of the federal poverty line, and free to those who are at or below 130 percent of the poverty line or who are homeless, in foster care, or in certain other programs. In 2012 in the lower 48 states and the District of Columbia, children living in a family of four on $40,000 or less a year would be eligible for reduced price meals; the free-lunch cut-off for the same family would be $30,000.
From a research and policy perspective, however, experts say food-aid eligibility gives an incomplete picture of the resources of students in poverty, and no information about students who don't qualify. Moreover, those poverty counts notoriously underrepresent students as they get older and more self-conscious about applying for free or reduced-price lunch.
"There are many problems regarding the use of free and reduced-cost lunch," said Henry M. Levin, a research panelist and an economics and education professor at Teachers College, Columbia University, who is now on sabbatical at Peking University in Beijing.
"It does not distinguish in a sensitive way differences along the entire spectrum of SES," he noted in an email to Education Week. "Even for the poor or relatively poor, there are large differences" within the range of free-lunch eligibility.
The governing board has tried in the past to fill in the gaps using the background questionnaire students complete along with NAEP, according to William Ward, a senior research scientist for assessment at NCES, which administers NAEP. But some of those questions have become outdated or have not been found to be relevant to a child's real socioeconomic status.
"We used to ask, 'Do you have a washer-dryer?' but now everyone has a washer-dryer," Mr. Cowan said. "We used to ask, 'Do you have a cellphone?' Now, do any of your students not have a cellphone?"
More Than Income
The updated measure of socioeconomic status will look at broader resources and learning supports, Mr. Cowan said.
It will start with the "big three": the family's income, parents' level of educational attainment, and whether and where they are employed. This year's administration of NAEP has also tried out new background questions, including how long the child has lived in the United States, how many family members live with the child, and how many adults in the home have a job.
Because elementary students in particular may have difficulty identifying these, the governing board is considering supplementing the data with information from the U.S. Census Bureau's American Community Survey, an annual study of a representative sample of 3.5 million households nationwide that asks about family structure, employment and income, transportation, and other details.
The NAEP student survey would still include questions about home possessions that research has shown to be related to student achievement, such as access to the Internet and the number of books in the home, Mr. Ward said. But the board is considering supplementing the "core" SES measures with other indicators of resources in the child's neighborhood and school that could highlight differences between students living at the same income level in different areas.
For example, an 8th grader in New York's Spanish Harlem neighborhood could still have access to libraries and museums, while a peer in rural southern Utah may have no local library but live a bike ride away from national parks.
Indicators of school and neighborhood supports also may be pulled from administrative data and from the Census Bureau, such as the degree of concentration of poverty or linguistic isolation, the average educational degree earned, and the employment levels in the neighborhood.
The governing board panel plans to present its proposed socioeconomic indicators at the annual meeting of the American Educational Research Association in April before piloting their use in 2014 and reporting the results in 2015.
Vol. 32, Issue 14, Pages 6-7Published in Print: December 12, 2012, as NAEP Seeks to Test New Measure of Student Poverty