School

State tests given from third through eighth grade can be placed on a single national scale, which puts districts and counties in different states side by side. This part asks how much of the difference in what children learn lines up with the district or county they live in, and how much lines up with family income and race inside the same place. Every comparison is between districts or counties. None describes an individual student, and an average score alone does not measure what a school contributes, because it reflects families, neighborhoods, health and schools together.

Chapter 2.1, Test scores, class and race measures how far apart county average scores are, in grade levels, and shows that average achievement moves with county income, within states as well as across them. It then turns to the gaps between groups inside the same districts. It compares white and Black students, and students who are and are not economically disadvantaged, and asks how much each gap varies from county to county. It relates the white-Black gap to school segregation and to the income gap, and it reports studies of Head Start and school spending under their own designs. The scores are state tests placed on the national scale and pooled over many school years, and small places are pulled toward the average, which narrows the spread the chapter reports.

Chapter 2.2, Gender gaps in reading and math compares girls and boys within the same districts, which holds schools and neighborhoods fixed in a way that a comparison of richer and poorer districts cannot. In reading, girls score higher across nearly all districts, and the sign is the same in each state, though the size differs by state. In math the gap has no common sign and moves with district income. A gap that keeps one sign in every kind of district does not depend on the district, and a gap whose sign changes with district income varies with place, so the chapter reads the two gaps differently. It does not test why the math gap moves with income.

Chapter 2.3, Learning growth and starting level (planned) will move from levels of achievement to growth, and ask whether students in districts that start behind learn faster or slower between third and eighth grade. The chapters above compare where scores stand at one time and say nothing about how fast children in any district learn.