From 2004 to 2007, gaps in percentages proficient between African American and white students narrowed in both reading and math at the elementary and middle grade levels analyzed. At the high school level, African American achievement gaps narrowed in math but widened in reading.

Similarly, gaps in percentages proficient between low-income and other students narrowed in both reading and math at nearly all grade levels analyzed. The exception was in high school reading, where the gap widened.

Gaps in percentages proficient between Latino and white students narrowed in both reading and math at the high school level.

Gaps in percentages proficient between Native American and white students narrowed in math but widened in reading at the high school level.

In general, the overall trends on state tests and NAEP moved in the same direction, though gains on NAEP tended to be smaller (NAEP tests are not aligned with any specific state’s academic standards). The most agreement was in fourth-grade mathematics. Of the 33 states with sufficient state test and NAEP data, 31 showed gains on both assessments.

Other explanations for gains

In an effort to explain why there were more gains than declines in performances, the CEP report noted “students may be learning more as a result of improvements in curriculum and instruction and extra interventions for urban students and for lower-performing students and subgroups.” The CEP listed these examples:

• Many school districts are devoting more instructional time to reading and math—often

by reducing time for other subjects.

• Students and teachers may have become more familiar with the format and content of

state tests as a result of test preparation activities.

• Narrow types of test preparation may have led to “score inflation” on state tests, meaning that scores have gone up on state tests without students having learned the broader knowledge of the subject being tested.

• Subtle manipulations in test design not otherwise addressed in this study could have

made some state tests easier over time. Examples include small changes in the choice of