Richard Allington, Anne McGill-Franzen, Gregory Camilli, Lunetta Williams, Jennifer Graff, Jacqueline Zeig, Courtney Zmach, Rhonda Nowak; University of Tennessee, KnoxvilleJournal of Reading PsychologySeptember 2010, forthcoming
Richard Allington, Anne McGill-Franzen, Gregory Camilli, Lunetta Williams, Jennifer Graff, Jacqueline Zeig, Courtney Zmach, Rhonda Nowak; University of Tennessee, Knoxville
Journal of Reading Psychology
September 2010, forthcoming
It’s well known that low-income kids lose academic ground from June to August, while middle- and upper-income kids are apt to continue to gain (or at least stay constant) due to camp, libraries, and other summer enrichment. The experimental study reported here sought to test one way of ameliorating summer learning loss. Researchers supplied students from seventeen high-poverty schools in Florida with a dozen self-selected books each summer for three years. The control group, kids drawn from the same schools and matched to the treatment population on a variety of demographic and academic variables, received no books. Three summers later, the treatment students outperformed the control group on Florida’s state reading test (FCAT)—and reading gains from the poorest kids were even larger. Not surprisingly, most youngsters chose books that pertained to pop culture—not “curriculum relevant” titles. Evidently reading about Britney Spears and Hannah Montana (a couple of the favorites) is better than no reading at all. Analysts estimated that supplying low-income pupils with summer reading costs about $50 per child, but that the overall impact of this intervention is similar to that of (the much more expensive) summer school. In other words, this learning loss intervention is cheaper than solutions currently being used, and gets a lot more bang for the buck. While it’s typically bad practice for researchers to both provide and evaluate an intervention, the study is nonetheless worth attention, especially in tight economic times when we’re looking for cost-effective measures to improve student achievement. Look for it in the fall edition of the Journal of Reading Psychology.
Joshua Breslau, Elizabeth Miller, W-J Joanie Chung, and Julie B. Schweitzer; UC Davis School of Medicine
Journal of Psychiatric Research
June 2010
It’s no surprise that psychiatric and substance-abuse disorders have a negative impact on graduation rates, but teasing out the origins of these disorders could help to target efforts at intervention. Researchers used data collected in 2001-02 on the National Epidemiological Survey of Alcohol and Related Conditions for nearly 30,000 students. After isolating the most significant indicators from commonly overlapping psychiatric and substance-abuse disorders, and adjusting for demographics and other variables, they found that one-third of students with the most common form (of three types) of attention-deficit hyperactivity disorder (ADHD) drop out or do not graduate on time. That is more than twice the percentage of students with no disorder, of whom 15.2 percent drop out. The surprise here is that ADHD appears to be more predictive of dropping out than conduct disorder, which is thought to be the most prognostic, and which refers to a group of behavioral and emotional problems that manifest in aggression and difficulty following directions (think: the student who is always acting out). But the third most predictive behavior/disorder is even more surprising: cigarette smoking. Twenty-nine percent of smokers drop out compared to drug users at 25 percent, drinkers at 20 percent, and a host of other disorders, such as mania, panic disorder, and post-traumatic stress disorder, in the high-teens/low-twenties. Perhaps none of this is shocking, but teasing out the oracular effect of individual disorders—often clumped together, but requiring very different intercessions at distinctive times in a student’s education—can only serve to make interventions that much more efficient, and hopefully, effective. Get it for a fee here.
U.S. Department of Education
August 2010
This week, ED launched a new online database that presents state-level achievement, graduation, accountability, demographic, and budgetary data in a single resource. Because it’s billed as the analyst’s dream come true, we decided to have a look ourselves. Here’s what we found: The data are not new—all of the information that the database compiles could previously be found scattered across ED’s website—but it does assemble that information and allow data to be compared across states and years. Sources include departments within ED, the National Center for Education Statistics, and the College Board, as well as self-reported state numbers. And it is indeed user-friendly: Drop-down menus make it easy to construct a table of a single state’s data or to compare data across multiple states. This allows for quick and straight-forward results, a handy tool for any researcher, journalist, or policy wonk (although ED cautions that state test scores should NOT be compared to each other or to NAEP). The database also includes some nifty bits, such as data on how many students took advantage of the opportunity to transfer from schools that did not meet adequate yearly progress under NCLB—and presents some data graphically. Sadly but unsurprisingly, given the data they had to work with, you won’t find anything here at the district or school level. And you need to be wary of self-reported state data, the more so when trying to compare states. Still, it’s a neat new tool, and a welcome shift towards transparency and accountability on the part of ED. Give it a test drive here.
In a guest editorial here last week, Sol Stern observed that New York State had significantly raised the bar for meeting proficiency requirements on state tests and that fewer students are meeting the new standard (“The testing mess,” August 5, 2010). Seeking to parlay that into a wholesale attack on New York City’s progress since Mayor Bloomberg took over in 2002, Stern and Diane Ravitch have launched several recent broadsides. The facts tell a different story, however.
It's no secret that many states have low standards; and the gap between “proficiency” on state tests and on the National Assessment of Educational Progress (NAEP) exams is typically very large. Mayor Bloomberg pointed that out in a 2006 op-ed with Jeb Bush, and recommended that the "well-respected NAEP…should become an official benchmark for evaluating states’ standards.” So NYS did the right thing in raising standards. But that doesn’t erase the fact that, by multiple measures, NYC has made substantial progress.
NAEP
Let’s start with NAEP, theldquo;gold standard.” In fourth grade, on the Mayor's watch, NYC has made big gains—11 scale-score points in English and 11 in math. The percentage of kids proficient in math went from 21 percent to 35 percent—a 67 percent increase—and the percentage proficient in English went from 19 percent to 29 percent—a 53 percent jump. Indeed, NYC’s performance now matches that of the entire nation in fourth grade, even though NYC serves a much more challenging population. That’s called “closing the achievement gap.”
In the eighth grade, the results are mixed and trending up. We’ve gained 7 points in math, mostly since 2005, and, while we’re flat in English, on the last exam we were up 3 points, boding well for the future. Overall, NYC’s gains on NAEP dwarfed those of rest of New York State and were also greater than those of the nation. Even Ms. Ravitch has acknowledged NYC's “significant progress.”
New York State tests
Although the state tests need to be improved, NYC's performance nevertheless reflects real progress. First, let’s look at how NYC would have performed in the past under the new, much more rigorous proficiency standards adopted by the state. We’ve had grades 3-8 testing in NYS since 2006 and, applying the new standards, our overall proficiency rate would have gone up by 6.4 points in English (from 36 to 42.4 percent) and by 22.1 points in math (from 31.9 to 54 percent). Going back to 2002, the gains are substantially larger.
Second, regardless of how you define proficiency, large increases in scale scores matter. Numerous longitudinal analyses in NYC and NYS have shown that students with a scale score of 690 in the eighth grade, for example, have a far greater likelihood of graduating and meeting NYS’s new definition of “college-ready” than do students with a 670. Similarly, students who score better on the state tests in the lower grades are more likely to score better in the higher grades.
NYC has repeatedly made those kinds of gains. From 2006-2010 our scale scores went up across the board in grades 3-8, by an average of 23 points in math and 13 points in English. Those are big gains and, if we go back to 2002, they’re considerably larger—46 points in fourth grade math, for example.
The dimension of these gains is readily seen by comparing NYC with two groups that took precisely the same tests: (1) other large urban school districts in NYS (called the “big four”: Syracuse, Yonkers, Buffalo, and Rochester); and (2) the rest of NYS districts. In 2002, NYC was close to the big four and distant from the rest of the state. Since then, we’ve moved away from the big four and much closer to the rest of the state. In fourth grade math, for example, we were 3 scale-score points ahead of the big four in 2002 and 22 scale-score points behind the rest of NYS; by 2010, we were 19 scale score points ahead of the big four and only 6 points behind the rest of NYS.
The same pattern holds for proficiency results, even with the declines on the latest tests. In eighth grade math, we were 10 percentage points ahead of the big four in 2002 and almost 30 percentage points behind the rest of NYS. Today, we’re 25 percentage points ahead of the big four and 15 behind the rest of NYS. On fourth grade English, we were 4 percentage points ahead of the big four in 2002 and 26 behind the rest of NYS. Now we’re 13 percentage points ahead of the big four and 19 behind the rest of NYS.
Several other key comparisons underscore this progress. NYS has sixty-two counties and, in 2002, NYC's five were at the bottom. Today, all five have made major gains, with two now in the top half. The same results are seen when the thirty-two school districts in NYC are compared to the thirty-two largest districts outside NYC. From 2002 to 2010, the city districts dramatically and consistently outgained the others. Finally, at the school level, in 2002, 9 percent of NYC’s schools were ranked in the top quartile statewide, while 62 percent were in the bottom. Today, the number in the top quartile has doubled to 18 percent, and the number in the bottom has decreased by almost one-third to 44 percent.
High school graduation
NYC's gains in graduation and college-going rates are also large. In the decade before the Mayor took over, the city’s graduation rate had stagnated in the mid-forties. Last year, it was 63 percent.
This increase has resulted in significantly more NYC students going to college. From 2002 to 2009, the number of graduates attending City University of New York (CUNY) colleges went from 16,000 to over 25,000—a 57 percent increase—while the number attending non-CUNY colleges also increased. At the same time, the percentage of students requiring remediation at CUNY decreased, meaning that more than 5,000 additional NYC students—an almost 80 percent increase—were college ready at CUNY in 2009 compared to in 2002.
F. Scott Fitzgerald once observed, “The true test of a first-rate mind is the ability to hold two contradictory ideas at the same time.” That’s what’s required to assess NYC’s record. On the one hand, NYC has a lot to do to improve student performance: More students need to graduate and they need to be college-ready. On the other hand, that recognition doesn’t remotely undermine the demonstrable progress NYC has made during the past eight years.
Klein is chancellor of the New York City Department of Education.
They say boys are from Mars, and girls are from Venus, and Imagine Southeast Public Charter School in D.C. couldn't agree more. Part of a growing experiment in single-sex education, Imagine is a "dual academy," which means it serves both boys and girls, but keeps them separated in different classrooms except for special occasions. Imagine’s is one of three popular single-sex models: serving just one gender, serving both genders in only single-sex classrooms, as Imagine does, or serving both but using a mix of co-ed and single-sex classrooms. The uptick in gender-specific education comes after a 2006 change to federal regulations made single-sex classrooms and schools easier to create. But the research is fuzzy. On the one hand, there's evidence that boys and girls not only have different learning styles—boys are competitive and can't sit still, while girls are collaborative and calm—but also biological differences. For example, girls need more time to complete an activity because stress decreases blood flow to a girl's brain, making her less ready to learn, whereas stress increases blood flow to a boy's brain, making him more alert. On the other, early childhood studies show that boys and girls have similar aptitude and preferences at birth, and it's cultural influences, such as that girls should play house and boys never cry, that account for much of the difference. Single-sex education arguably exacerbates these. Then there's the practical side: Wiggling boys distract girls, while girls' ability to read sooner frustrates boys and may make them disengage. Doesn't seem like there's much agreement, but this is certainly an interesting development on which it's worth keeping an eye.
“Separate but equal: More schools are dividing classes by gender,” by Karen Houppert, Washington Post Magazine, August 8, 2010
Turns out Lehman Brothers and the rest of Wall Street weren’t the only ones to make risky and complex financial deals in the waning days of the credit bubble. Denver Public Schools entered into one such arrangement in April 2008 to close a $400 million gap in its pension obligations to teachers. Instead of issuing regular bonds, the district offered certificates with variable interest rates and an interest-rate swap. Then interest rates collapsed, the market for Denver's debt dried up, and now DPS is out almost as much cash as the original loan, plus another $115 million in interest and fees. It’s not alone: Los Angeles’ municipal government and a few school districts in Pennsylvania are trying to renegotiate or unwind similar interest-rate deals. But there’s more to this story than just a few bad decisions: What made these deals necessary. DPS and other districts face gargantuan pension obligations to teachers, billions more than they can actually pay. Yet these unaffordable promises are now locked into state law, or even state constitutions. Even Colorado’s momentous move earlier this year to reduce yearly automatic raises on pensions, the only such to affect current, as opposed to future, retirees, is just a drop in the bucket. Though CO’s pension fund has avoided insolvency—when “no one would have been paid anything,” as pension fund head Meredith Williams dryly points out—at least for now, it doesn’t solve the larger structural issue of government pension funds more generally. This financial crisis calls for a radical makeover.
“Exotic deals put Denver Schools Deeper in Debt,” by Gretchen Morgenson, New York Times, August 5, 2010
“Battle Looms Over Huge Costs of Public Pensions,” by Ron Lieber, New York Times, August 6, 2010
Joshua Breslau, Elizabeth Miller, W-J Joanie Chung, and Julie B. Schweitzer; UC Davis School of Medicine
Journal of Psychiatric Research
June 2010
It’s no surprise that psychiatric and substance-abuse disorders have a negative impact on graduation rates, but teasing out the origins of these disorders could help to target efforts at intervention. Researchers used data collected in 2001-02 on the National Epidemiological Survey of Alcohol and Related Conditions for nearly 30,000 students. After isolating the most significant indicators from commonly overlapping psychiatric and substance-abuse disorders, and adjusting for demographics and other variables, they found that one-third of students with the most common form (of three types) of attention-deficit hyperactivity disorder (ADHD) drop out or do not graduate on time. That is more than twice the percentage of students with no disorder, of whom 15.2 percent drop out. The surprise here is that ADHD appears to be more predictive of dropping out than conduct disorder, which is thought to be the most prognostic, and which refers to a group of behavioral and emotional problems that manifest in aggression and difficulty following directions (think: the student who is always acting out). But the third most predictive behavior/disorder is even more surprising: cigarette smoking. Twenty-nine percent of smokers drop out compared to drug users at 25 percent, drinkers at 20 percent, and a host of other disorders, such as mania, panic disorder, and post-traumatic stress disorder, in the high-teens/low-twenties. Perhaps none of this is shocking, but teasing out the oracular effect of individual disorders—often clumped together, but requiring very different intercessions at distinctive times in a student’s education—can only serve to make interventions that much more efficient, and hopefully, effective. Get it for a fee here.
Richard Allington, Anne McGill-Franzen, Gregory Camilli, Lunetta Williams, Jennifer Graff, Jacqueline Zeig, Courtney Zmach, Rhonda Nowak; University of Tennessee, Knoxville
Journal of Reading Psychology
September 2010, forthcoming
It’s well known that low-income kids lose academic ground from June to August, while middle- and upper-income kids are apt to continue to gain (or at least stay constant) due to camp, libraries, and other summer enrichment. The experimental study reported here sought to test one way of ameliorating summer learning loss. Researchers supplied students from seventeen high-poverty schools in Florida with a dozen self-selected books each summer for three years. The control group, kids drawn from the same schools and matched to the treatment population on a variety of demographic and academic variables, received no books. Three summers later, the treatment students outperformed the control group on Florida’s state reading test (FCAT)—and reading gains from the poorest kids were even larger. Not surprisingly, most youngsters chose books that pertained to pop culture—not “curriculum relevant” titles. Evidently reading about Britney Spears and Hannah Montana (a couple of the favorites) is better than no reading at all. Analysts estimated that supplying low-income pupils with summer reading costs about $50 per child, but that the overall impact of this intervention is similar to that of (the much more expensive) summer school. In other words, this learning loss intervention is cheaper than solutions currently being used, and gets a lot more bang for the buck. While it’s typically bad practice for researchers to both provide and evaluate an intervention, the study is nonetheless worth attention, especially in tight economic times when we’re looking for cost-effective measures to improve student achievement. Look for it in the fall edition of the Journal of Reading Psychology.
U.S. Department of Education
August 2010
This week, ED launched a new online database that presents state-level achievement, graduation, accountability, demographic, and budgetary data in a single resource. Because it’s billed as the analyst’s dream come true, we decided to have a look ourselves. Here’s what we found: The data are not new—all of the information that the database compiles could previously be found scattered across ED’s website—but it does assemble that information and allow data to be compared across states and years. Sources include departments within ED, the National Center for Education Statistics, and the College Board, as well as self-reported state numbers. And it is indeed user-friendly: Drop-down menus make it easy to construct a table of a single state’s data or to compare data across multiple states. This allows for quick and straight-forward results, a handy tool for any researcher, journalist, or policy wonk (although ED cautions that state test scores should NOT be compared to each other or to NAEP). The database also includes some nifty bits, such as data on how many students took advantage of the opportunity to transfer from schools that did not meet adequate yearly progress under NCLB—and presents some data graphically. Sadly but unsurprisingly, given the data they had to work with, you won’t find anything here at the district or school level. And you need to be wary of self-reported state data, the more so when trying to compare states. Still, it’s a neat new tool, and a welcome shift towards transparency and accountability on the part of ED. Give it a test drive here.
