Schools across the nation are held accountable under the No Child Left Behind Act (NCLB) for raising student achievement in math and reading. But each state gets to be its own policeman by designing the tests its students take and determining the scores that count as success. The National Assessment of Educational Progress (NAEP), administered every two years in grades 4 and 8 to representative samples of students from each state, serves as a check on what the states report using their own assessments. It also provides a measure of trend over time, something that is often lacking in the states’ own data because of frequent changes in state assessments.
Much is typically made in policy circles of the contrast between the percentages of students reported to be proficient on NAEP vs. on state tests. For example, NAEP rates 28 percent of Tennessee’s eighth graders as proficient in reading whereas the state of Tennessee finds more than 90 percent of its students to be proficient. Such disparities fuel the drive for common standards as embodied in the Common Core effort led by the National Governors Association and supported by Race to the Top funding from Washington. However, both the state and the NAEP standards for proficiency are arbitrary. It is far more informative to examine scale scores and ask if there has been progress. While state tests and NAEP convey very different messages with regard to how many students are deemed proficient, they are largely in sync in marking whether scores have moved up or down.
NAEP reading scale scores are up in 2009 at eighth grade compared with 2007, and unchanged at fourth grade. This mirrors the results for mathematics that were released in December. Historically gains have been more likely and larger at fourth grade than at eighth grade, so the shift of the action to eighth grade is interesting. It probably reflects children who benefited from better instruction in elementary school a few years ago now being tested in eighth grade, or education reforms that were focused on elementary school moving into middle school, or both.
The magnitude of the overall improvement in scale scores since 2007 is small: eighth grade scores were 1 point higher than in 2007. However, this represents about a month of additional schooling, which is not trivial. Gains were twice as large for the lowest performing students.
When a two-year testing cycle for NAEP was introduced in 2003, replacing a roughly four-year cycle, it became less important to look at the changes from one testing to the next, and more important to examine longer trends. Through this lens there is more to celebrate. Since 2002, reading scores for Black students at fourth grade are up 6 points, equivalent to about two thirds of a year of schooling. They’re up 5 points for Hispanics. At eighth grade the gains have been smaller but real for these two groups, 1 point for Blacks and 2 points for Hispanics. Gains over this time frame have been much larger for mathematics, e.g., 9 points for Blacks at eighth grade.
Commentators on NAEP, from cabinet secretaries to casual bloggers, can’t seem to resist the temptation to opine on whether the results mean that NCLB is working. With the Obama administration’s blueprint for reauthorization just released, the appeal of this line of speculation will be overwhelming. In fact, NAEP isn’t up to that task. Many things have changed during the time frame of NCLB, including a dramatic shift in the demographics of the school aged population, the growth of charter schools and home schooling, and policies for including students with disabilities and limited English proficiency in assessments. And let’s not forget changes in the culture as a whole that affect academic performance such as the growth of use of the Internet and digital media by school-aged children and the reduction in the teenage pregnancy rate. Besides, NCLB and the administration’s blueprint aren’t single things. They’re packages of many different policies and programs. Forget NAEP and dig into the research on those individual policies and programs if you want insight on what works.