If the No Child Left Behind Act is raising student achievement, it鈥檚 not conclusive based on the latest results from the National Assessment of 91制片厂视频al Progress: Reading achievement has remained essentially flat over the past two years, while gains in math have slowed.
In reading, the average national scores have risen just a point for 4th graders since 2003鈥攖o 219 on a 500-point scale鈥攁nd dropped a point for 8th graders, to 262, according to the 2005 NAEP results, released here Oct. 19. In fact, reading scores for both grades have stayed about the same since 1992, the first year data were available.
Math scores, which have climbed dramatically since 1990, when students first took the current version of the NAEP test in that subject, rose again in 2005, but more modestly this time. Fourth graders have posted a 3-point gain since 2003, reaching a national average of 238, while the average score for 8th graders has risen a point, to 279. Some 36 percent of 4th graders and 30 percent of 8th graders were rated 鈥減rofi-cient鈥� on the 2005 math test. Larger proportions than in the past also demonstrated at least basic skills: 80 percent of 4th graders and 69 percent of 8th graders.
In July, U.S. Secretary of 91制片厂视频 Margaret S. Spellings pointed to gains on what is called the long-term-trend NAEP as 鈥減roof that No Child Left Behind is working.鈥� That exam, based on a much smaller sample of youngsters nationwide, showed a 9-point gain in math scores and a 7-point gain in reading scores for 9-year-olds, as well as a 5-point gain in math scores for 13-year-olds, from 1999 to 2004. Ms. Spellings said the Oct. 19 results 鈥渃onfirm that we are on the right track with No Child Left Behind.鈥�
The newly released results, for a much larger sample of students鈥�660,000 nation-wide鈥攁lso show an upward trend since 2000 in both mathematics and reading in grade 4 and in math in grade 8. But they also show that progress has slowed or stagnated in the past two or three years, when the No Child Left Behind law might have been expected to have an impact.
The federal measure, which President Bush signed into law in January 2002, is designed to drive broad gains in student achievement, with particular emphasis on reading and math. The NAEP tests provide what experts consider the most reliable national picture of student achievement in core subjects. The national assessment also provides state-level results.
Minority Gains
In a telephone briefing for reporters, Secretary Spellings highlighted as particularly encouraging a 2-point gain in reading among African-American 4th graders between 2003 and 2005, from 198 to 200, and a 3-point gain among Hispanic 4th graders, from 200 to 203. 鈥淚鈥檓 very pleased that we鈥檙e making steady progress, which is driven by increased progress by Hispanic and African-American students, which is the whole point of No Child Left Behind,鈥� she said.
Just how much to celebrate a few points鈥� gain, however, is open to debate. While the differences in test results between generally lower-scoring black and Hispanic students and their generally higher-scoring white peers narrowed by a point or 2 from 2003 to this year, depending on the grade and subject, the minority students are still more than 20 points behind. Though the new NAEP math results are 鈥渁n affirmation that much of what we are doing is working,鈥� said Cathy L. Seeley, the president of the National Council of Teachers of Mathematics, 鈥渢he results for minority students and children in poverty remain unacceptably low.鈥�
At a Washington press conference held to release the results, Grover J. Whitehurst, the acting commissioner of the National Center for 91制片厂视频 Statistics, the arm of the U.S. Department of 91制片厂视频 that oversees NAEP, said: 鈥淚f we could eke out a 2-point change each year [for minority youngsters] over a 15-year period, the gap that existed at the beginning of that period would essentially be erased.鈥�
Middle School Dip
Yet based on the current rate of improvement on NAEP, the No Child Left Behind Act鈥檚 goal that all children will be proficient on state tests in math and reading by 2013-14 appears unlikely. Even assuming that what counts for 鈥減roficient鈥� in many states is closer to the 鈥渂asic鈥� achievement level on NAEP, the proportion of students at or above basic budged by only a few percentage points between 2003 and 2005 at most. Currently, 36 percent of 4th graders and 27 percent of 8th graders still score below basic in reading.
鈥淚 think it does show us that we鈥檙e going to need to accelerate our progress at all grade levels and with all kids if we鈥檙e to meet those goals,鈥� said Ms. Spellings in the Oct. 19 telephone briefing with the news media after the release. 鈥淲e actually won鈥檛 do it, if we don鈥檛 think we can do it.鈥�
She also argued that the federal law has yet to be implemented fully; for example, states must test students annually in reading and math in grades 3-8 and at least once in high school starting this school year. And the law鈥檚 Reading First investments in grades K-2 will take time to affect results in the upper grades, the secretary said.
Timothy Shanahan, a professor of education at the University of Illinois at Chicago and the president-elect of the International Reading Association, agreed that Reading First hits a relatively small proportion of schools, primarily those that are low-income and low-achieving. The $1 billion-a-year program is supposed to support enhanced research-based reading instruction. 鈥淯ntil we actually get some data out of those schools, I think you鈥檇 have to say that the jury is still out on Reading First,鈥� Mr. Shanahan said.
鈥淭he broader question of NCLB is a little different because that is clearly something that鈥檚 supposed to be hitting everybody,鈥� he added. 鈥淐learly, we should start to see some impact of that; the issue is how quickly should you see it?
鈥淚f you just look at the overall scores, particularly at the 4th grade level,鈥� Mr. Shanahan continued, 鈥渢here鈥檚 absolutely no doubt that the scores are up a bit, but it鈥檚 a small bit.鈥�
But Darvin M. Winick, the chairman of the National Assessment Governing Board, which sets policy for the congressionally mandated NAEP, said, 鈥淭he expectation for rapid change in a large, complex organization such as public education probably should be moderate.鈥�
鈥淎nticipating in two-year intervals a lot of change is probably unrealistic,鈥� he said at the press conference.
While the nation might be on a path of elementary school improvement, the reading performance of 8th graders was clearly troubling, however. 鈥淎s a former English teacher, I am distressed by what I see in reading,鈥� said John H. Stevens, the chairman of the governing board鈥檚 reporting and dissemination committee. 鈥淚 think the situation is we stop working explicitly on reading skills and start focusing on content鈥� after elementary school, he said in an interview. Schools need to work more explicitly on reading skills all the way through high school and across the disciplines, Mr. Stevens argued.
Mr. Shanahan also said efforts can鈥檛 stop with the early grades. 鈥淚f you only focus your reform attention in reading on the lower grade levels, that will not necessarily translate into higher achievement across the board.鈥�
鈥淚 wish I could just blame the policymakers,鈥� he continued. 鈥淭here are a lot of people in the field who pushed very hard for improvements in the primary grades with the assumption that would just fix it.鈥�
State Differences
On the 2005 assessment, the state-level results were mixed, but several states showed some returns on their investment in improved instruction and increased accountability measures, according to Mr. Winick of the NAEP governing board.
鈥淚f you look at the state-by-state data, the states that are pressing for results tend to get them,鈥� he said. 鈥淓very time you look, they have more dramatic gains in math and reading.鈥�
That was the case particularly at the 4th grade level, where Texas saw a 4-point gain in reading and a 5-point increase in math. Idaho and Pennsylvania also achieved higher scores in that grade for both subjects. In all, seven states attained statistically significant increases in reading scores among 4th graders, while seven saw their scores drop among 8th graders. On the math test, more than 30 states, including Texas, showed some improvement on the 4th grade test, while seven did so for 8th graders.
This year, for the first time, the NAEP reports also include information on the changing student racial and ethnic composition of states between 1992 and 2005. 鈥淎s our country becomes increasingly diverse, so, too, do students populations, which vary both between states and over time,鈥� NAGB鈥檚 Mr. Stevens said. 鈥淎s a result, achievement data and trends should be interpreted as a reflection of changing demographics as well as school effectiveness.鈥�
As an example, he noted in the past 15 years, the size of the Hispanic student population has more than doubled nationally. 鈥淐omparisons without that kind of analysis are mostly unfair and really do not reveal what鈥檚 going on,鈥� Mr. Stevens said.
鈥淚t鈥檚 notable that as our student population has become increasingly diverse, the scores have continued to rise,鈥� Secretary Spellings argued.
The national assessment, which was first administered in 1969, also has become more inclusive over time, offering some accommodations to students with disabilities or limited English fluency who need them. That has led to questions about how to interpret the results, particularly when the percent of such students excluded from the exams varies across states.
Delaware, Louisiana, and Virginia, for instance, excluded 10 percent or more of their 4th graders with disabilities from the 2005 reading exam, while New Mexico and Texas excluded 7 percent and 6 percent, respectively, of their students still learning English.
While previous analyses have found only a 鈥渟light statistical relationship between different exclusion rates across states and scores at the aggregate level鈥� in any given year, said Mr. Whitehurst of the Department of 91制片厂视频, 鈥渟tates who raise their exclusion rates tend to have scores that go up on the next assessment, and states who lower their exclusion rates tend to have scores that go down.鈥�
