Dayton Daily News education reporter Scott Elliott writes about schools, kids, teaching and learning.
Home > Blogs > Get on the Bus > Archives > 2006 > September > 05
Tuesday, September 5, 2006
Carnival entries due!
Entries for the 83rd Carnival Of Education, to be hosted in this space tomorrow, are due TODAY.
Submissions should be received no later than 9 p.m. (Eastern), 6 p.m. (Pacific). Contributions should include your site’s name, the title of the post, and the post’s URL if possible. If you have an education blog and would like to offer a post for inclusion in the carnival, email me at scemel@aol.com or use this handy submission form.
View last week’s edition, here and the Carnival’s archives over there.
Barring unforeseen circumstances, the exhibits should open Wednesday.
Permalink | | Categories: The Carnival of Education
Learning from Piqua
If there’s one thing that struck me most while reporting today’s story about income and test scores, it was just-retired Piqua Superintendent Jerry Clark talking about how common it was for him to see kindergarenters who didn’t just have the usual problems of not knowing their letters and numbers.
These kids had never seen a crayon before. They had never held a book or been read to.
In essence, they were starting five years behind other kids sitting in the same classroom. Now that’s a challenge for a teacher and a school.
Today’s story had its roots in some musings here at Get on the Bus not long ago. That got me wondering just how strongly income and test scores were connected in Ohio and what that said about the fairness of our state’s system of evaluating schools.
Now we know the connection is very strong, and that the state is looking for new ways to evaluate schools.
I didn’t get as much from Clark or about Piqua into the story as I would have liked, but this small, poor city in a mostly rural corner of Miami County is a good example of how hard the work can be for schools to make a noticeable difference when it comes to state report cards, even when they are helping kids learn.
I’ve been interviewing Clark for about a decade, talking to him about the challenges of educating kids in cities like Piqua. This is a town with a lot of proud, hard working people who care about their kids. But some of the poverty there is extreme. For the schools, change took time, it took focus, it took community support and a lot of hard work. Finally, this year’s state report card ranked Piqua “effective” and when you look at the rest of the state you find very few examples of a school district ranked as low as Piqua for median income with an effective rating.
But the frustration for Clark was holding together his base of support year-after-year and keeping people believing that the road they were on was the right one. People are understandably impatient when it comes to their kids and the quality of their education and every year report cards came out saying Piqua was among the worst around.
Ohio’s system set the bar, for most tests, at 75 percent passing. That was a long road for Piqua. For years, Clark was looking at data and other results that told him Piqua was improving, but it’s hard to keep people believing when the state keeps ranking the district near the bottom. It made it tougher to pass levies and occasionally stirred political unrest for school leaders.
In recent years, Ohio has moved toward a middle ground. Those 75 percent passing rates are still required but districts now can get a nudge up the rating scale if they show strong improvement. Right now the “growth” measure the state uses is not very sophisticated. Essentially it looks at how well districts move their average scores up.
With next year’s new value-added system, Ohio hopes to track the test score changes of individual kids and quantify growth by student, classroom, school building and school district. It’s a very sophisticated system that some experts are not even sure will work. Clark, for one, is hoping it does.
One of the experts I spoke to for the story told me that for an individual student the “who they are” influence of their family background usually accounts for 30 to 40 percent of their individual test score results. That means the teacher generallly can influence up to 70 percent of their test result through “what they learned.”
But when you aggregate individual students together in large groups it becomes much harder to move the overall passing rates. The “who they are” side of the coin influences the overall percentages much more strongly. The education holy grail is finding a way to measure “what they learned” for large groups on a wide scale.
Permalink | Comments (2) | Categories: Testing
Income, test scores strongly linked in Ohio schools
By Scott Elliott
Dayton Daily News
Sometimes turkey and trimmings can tell you nearly as much about a school as test scores.
Kettering’s Oakview Elementary School would seem like a good bet for high scores. Located just a mile down Ackerman Road from upscale Lincoln Park, it draws from a bustling suburban neighborhood in a good performing school district.
But come Thanksgiving, former principal Ron Sinclair saw another side — some families were not as well off as the neighborhood might suggest.
Before retiring last year, he helped deliver Thanksgiving meals to about 30 families for a school with 385 enrolled.
“We were surprised by how many names we got,” he said.
State data lists a quarter of Oakview’s students as “economically disadvantaged,” fourth most in the district, which may help explain why it has rarely ranked higher than fourth among the eight Kettering elementary schools for test performance.
That family income is related to test performance is no secret. But a Dayton Daily News computer analysis shows a particularly strong connection in Ohio. The median income for its 610 school districts strongly correlates to their “performance index scores,” a state measure of test performance across all grades, the analysis showed.
The test score connection with income is more than twice as strong as with other state report card factors like race, teacher pay, teacher education or school district spending.
While some experts question if the state relies too heavily on test scores to judge schools, Ohio already is moving toward an ambitious new approach that next year will seek to measure the “value added” by a school independent of other factors.
Come test time, Sinclair knew the number of needy kids would work against Oakview, even with a top-rate staff, parents who cared and solid resources, including a counselor and extra staff to make quick plans to catch up struggling kids.
“You can work smarter and collect data and analyze things, but there are some overriding variable you just can’t make up through hard work,” he said. “Sometimes you can do an extremely good job and it doesn’t show up in the test scores.”
Kim Kappler’s last job change took her from one of Ohio’ poorest school districts to one of its wealthiest.
In 2005, Kappler came from Norwood schools — ranked 505th out of 610 for median income in 2004 — to Oakwood, ranked eighth in the state on the same list.
“In both of those situations, parents love their kids and want what’s best for them,” said Kappler, Oakwood’s director of curriculum, instruction and testing. “But in Oakwood, the parents can provide things Norwood parents cannot.�?
And it’s not just summer camps and lots of books. Kappler’s list of advantages starts with the basics — eye and ear screenings, dental hygiene, good nutrition.
“Parents in Norwood just had less education and experience,” she said. “In Oakwood, parents recognize their success is based on their own educational attainment.”
Just as educators say household wealth, or lack of it, can reliably help or handicap a student’s school success, median income in a community powerfully predicts standardized test success for school districts.
To determine just how strongly test scores and income are connected, the Dayton Daily News compared the statistical relationship between 2004 median income from tax returns with just released performance index scoresfor all 610 Ohio school districts. The correlation was robust — more than twice what researchers expect for a strong connection.
When the same calculation was run for other factors on Ohio’s state report card — race, teacher pay, teacher training and school district spending and size — the connection was less than half as strong as for income.
An identical analysis for just the 82 Dayton-area districts gave the same result — income was by far the strongest predictor of test success.
Oakwood’s income, for instance, would predict high-scoring success, and yearly it is among the highest-rated districts in the state and locally. Northridge, the lowest-income area district, would shock statisticians if it ever reached a top test ranking. It never has.
Ohio’s system of evaluating schools — district report cards — relies heavily on tests to judge school district effectiveness. But some educators and researchers question if that approach is fair.
“All the talk of school failure, but it’s always the schools in the more affluent areas that are doing pretty good,” said Chris Lubienski, a University of Illinois researcher who studies the effect of student characteristics on test performance. “Schools in high poverty areas are always the ones in trouble.”
Count Ohio lawmakers among those interested in other ways to measure school district effectiveness. In 2003, the legislature required the state education department to put in place a “value-added” system that will launch with the 2007 report card. Next year, this system will use a complex statistical model, similar to one developed for the state of Tennessee by statistician Bill Sanders, to try to quantify teacher effectiveness by controlling for the effect of income and other factors.
Mitchell Chester, the assistant state superintendent who oversees Ohio’s testing program, said the state’s own analysis of the relationship between median income and test performance also found a strong connection, although the state found the correlation was not as strong when about 60 districts with very high median income over $40,000 were removed.
Chester cautioned the studies don’t prove wealth, or lack of it, predetermine test scores. Schools do make a difference, he said.
“Is in fact what is going on here is some school districts are providing a rich, engaging and intellectually challenging curriculum and some are not?” he said. “You tend to find in higher wealth communities there is a demand for that. In communities with lower wealth you are not as sure to find that curriculum demand.”
Quality teaching and good curriculum can move students to higher test scores, he said. The state must hold schools to its expectations for test performance, but the value-added system can help further identify which schools are doing the job of helping kids learn.
“One of the appeals of value-added is it’s a measure not of simple attainment but of whether school districts are moving the kids from where they started,” he said. “It will be interesting to watch the data unfold. Are the strongest gaining districts some of those with lower results when you look at achievement at a point in time?”
Lubienski agrees that schools do make a difference. Researchers commonly find teachers and schools can account for up to 30 to 40 percent of a school’s test result. But that still means 60 to 70 percent of a school’s scores are connected more strongly to outside factors, family income especially.
“We’ve known since the 1960s that despite what we’d like to think, schools don’t have as much of an effect on achievement as some of the background factors do,” he said. “But schools are the one area you can influence through public policy. You can’t legislate that parents read to their kids.”
Even so, Lubienski is not sold on Sanders’ value-added system. The specifics of the math behind his calculations have not been made public — Sanders’ methodology is proprietary. The system seeks to hold constant all outside factors that might influence test scores to examine just the school’s effect on each individual student.
Lubienski is skeptical.
“I just don’t think social science is at that point,” he said.
But Jerry Clark, Piqua’s just-retired superintendent, said years of watching kindergarteners hold their first crayons or open their first books led him to fight for new state measures of student growth.
“The rating system was designed only for one snapshot, as if everybody were playing with the same set of circumstances,” he said.
Using a wide spectrum of strategies, Piqua this year earned an “effective” state rating this year despite being one among the Dayton area’s poorest districts. Piqua’s test performance ranks well above where its median income would predict.
“There are school districts that are doing a great job that may never score as well as districts serving kids from more advantaged families,” he said. “I’m thrilled there will finally be some recognition.”
Permalink | Comments (2) | Categories: My Favorite DDN Stories, Testing
