|
|
9486 in the collection
Reading, Writing and. . . Statistics
Ohanian Comment: Here are two items from the same issue of the Wall Street Journal.
I wonder if any expert who is quoted saying the NAEP is "highly regarded" has ever read the reader comments on why they gave the points they did for open ended student answers. My analysis of NAEP reading passages provided inside dope on media headlines screaming, "NAEP Finds 71% of 4th graders score below the proficient level." It has been singularly ignored by the world. This information is important because corporate politicos are pushing for NAEP to become the national test.
by Carl Bialik
To understand the latest U.S. reading scores, it helps to have an education in statistics.
While fourth and eighth graders in urban public schools lagged behind national averages, a quick read of National Assessment of Educational Progress charts might suggest that city students still were showing promising gains last year compared with 2007. After all, in 10 of 11 cities the average fourth-grade score rose, as it did for eighth graders in six cities.
But the quick read might not accurately reflect the students' progress. Reading scores are like political polls. Only a sample of eligible American voters, meant to stand in for the electorate, are asked which candidate they would vote for. Similarly, only some fourth graders—2,400 or fewer in each of the 11 cities —were tested on their ability to perform tasks such as reading about bees and answering how they are important to plants or animals.
Choosing a sample raises the risk that the representatives aren't representative, and introduces a margin of error. In a quickly changing presidential race, shifts in public opinion can be drastic enough to overcome statistical uncertainty.
But changes in student reading aren't as extreme. So statisticians can't say whether a rise in reading scores to 213 from 210 among San Diego fourth-graders was real or just the result of statistical noise—and the same goes for increases in five other cities, and among urban students overall.
This same ambiguity accompanied the release of 2007 scores. Fourth-graders in eight cities showed gains, but just two of those were significant. Shifts in four of the eight cities to show gains among eight-graders also were washed out by the margin of error, which varies according to the sample size.
Bundling students into larger groups can add statistical power, as can measuring trends over longer periods of time. This is evident in the Trends in International Mathematics and Science Study test, the one used to compare American students with their peers world-wide.
U.S. eighth grade students' average math score in 2007, the latest available, wasn't significantly better than in 2003 or 1999, but their gains from 1995 were meaningful. Even so, Timss scores can't indicate whether American eighth graders did better or worse in 2007 than their counterparts in, say, Hungary, England, Russia, Lithuania or the Czech Republic.
Some trends simply are too gradual to pinpoint in the short term. The U.S. made progress in math from 1995 to 2007, but it "was just creeping," said Ina Mullis, a Boston College professor who oversees the Timss test. "It's actually pretty difficult for a huge educational system like our country's to make change sometimes."
Write to Carl Bialik at numbersguy@wsj.com
Literacy Scores Stall in Inner Cities
By Stephanie Banchero
It's always interesting to see who gets quoted--and who doesn't.
Students in large U.S. inner cities are struggling to improve their reading ability, especially at middle-school levels, according to results from a national reading test released Thursday.
Only Atlanta and Los Angeles, two of the 11 urban centers that took the reading exam, showed statistically significant growth in eighth-grade reading from 2007 and 2009. They also were the only two to show growth since 2002.
Four districts notched gains at the fourth-grade level since 2007, while five showed progress since 2002.
The test scores are part of the urban National Assessment of Educational Progress, or NAEP. Known as the "nation's report card," the exam has been given in urban districts biennially to students in the two benchmark elementary and middle school grades.
The lackluster reading results follow the December release of urban math scores, which also showed stalled progress.
The stagnating scores of urban schools come at a key time in education reform. President Barack Obama has centered his education agenda on lifting the achievement of inner-city children. Their progress, especially in literacy skills, is critical to his push to keep the nation competitive in a global economy.
Education Secretary Arne Duncan lamented the middle-school scores during an interview Thursday. He said his administration wants districts to build more comprehensive reading programs that reach into middle and high schools.
"We tend to focus on the early grades, which is important," he said. "But if we want to really move the needle on reading, we need to tackle the tough reading challenges in middle schools."
Mr. Duncan has proposed a new $450 million literacy program for 2011 that would provide competitive grants to states that craft comprehensive literacy initiatives from pre-K through 12th grade.
School districts in six areas, including Chicago, New York and Washington D.C., have voluntarily participated since 2003. Seven districts took the test for the first time last year. A companion NAEP exam is given to students across the country, allowing a comparison between national and urban scores.
The highly regarded test is scored on a zero-to-500 point scale, with scores broken into "below basic," "basic," "proficient" and "advanced." Students are considered to have passed the exam at the proficient level.
The test is more rigorous than most state exams. So while many urban districts report impressive gains on state tests, they cannot claim such progress on the NAEP.
Michael Casserly, executive director of the Council of Great City Schools, said that trend data from 2003 show urban schools have made some strides in reading. But he said junior-high literacy continues to present a challenge.
"It appears students simply are not coming into middle school with a high enough level of comprehension skills," said Mr. Casserly, whose nonpartisan group represents 66 city districts. "This is not just in the cities, but a nationwide problem."
Overall, urban schools scored far below the national averages in fourth- and eighth-grade reading. Most urban districts posted higher scores on both exams. But in many cases, the gains were not statistically significant, meaning they could have occurred by chance and might not reflect real growth.
At fourth grade, four of 11 inner-city districts saw an uptick in performance between 2007 and 2009, with the District of Columbia and Houston notching the largest gains. D.C. jumped from an average score of 197 to 203. But even that kept them at "below basic," the lowest quartile on the scoring chart.
Houston improved from 206 to 211, landing them one step up in the "basic" category. The national average was 220 last year, the same score as 2007.
But eighth-grade reading proved a bigger hurdle. Only two districts showed progress, and even then, their scores were far below the nation. Atlanta posted a 250, Los Angeles a 244. The national average was 262. All of these scores are below the proficient level.
D.C. posted a slight gain if charter schools are removed from the equation. Unlike most national school districts, the D.C. charter schools are considered a separate entity from the D.C. public school system. In the national data, D.C. charter schools were included in 2007, but not in 2009, inflating the scores.
Catherine Snow, a reading expert at the Harvard Graduate School of Education, said the nation spent seven years under No Child Left Behind focusing on the basic reading skills of children.
Literacy efforts in many urban districts were driven by Reading First, President George Bush's $6 billion reading initiative for kindergarten through third grade.
"Reading First was a good effort at solving a problem that was defined too simply," she said. "We spent a lot of money focused on phonics and fluency and those are very important. But we have not spent enough time teaching kids comprehension skills and we are suffering because of it."
— Carl Bialik and Stephanie Banchero
Wall Street Journal
2010-05-21
http://online.wsj.com/article/SB10001424052748704513104575256763373287690.html
INDEX OF OUTRAGES
Pages: 380
[1] 2 3 4 5 6 Next >> Last >>
|
