Network News

X My Profile
View More Activity

Posted at 6:30 AM ET, 07/22/2010

Poverty, student achievement, HCZ: Berliner

By Valerie Strauss

Alexander Russo and Jay Mathews got into it yesterday here over the Harlem Children’s Zone and a new paper questioning the effectiveness on student achievement of the zone’s wrap-around social and education services. Mathews defended the zone and its charter schools, Russo went after Mathews, and the two made up and carried on to fight another day.

But there are unanswered questions about the paper that was published by the Brookings Institution and written by Senior Fellow Grover J. “Russ” Whitehurst and Research Analyst Michelle Croft.

Whitehurst and Croft said that the zone's myriad health and other services for children who attend its charter schools didn't raise test scores as high as some other New York City charters that did not offer the same out-of-school programs.

“There is no compelling evidence that investments in parenting classes, health services, nutritional programs and community improvement in general have appreciable effects on student achievement in schools in the U.S.,” it says. “Indeed, there is considerable evidence in addition to the results from the present study that questions the return on such investments for academic achievement."

This has more than academic interest because the zone is being used as a model by President Obama in his Promise Neighborhood's Initiative that calls for the creation of similar communities in 20 cities.

I sought a reaction from David Berliner, Regent's professor emeritus at Arizona State University, a prominent researcher and educational psychologist who has studied the connection between poverty and student achievement extensively. Here’s what he told me. It’s long but makes interesting reading:

Hi Valerie, I think he [Whitehurst] is dead wrong, but he is a lover of randomized clinical trails and if that is the criteria for evidence that he uses he is more likely to be right than wrong. But there is lots of other evidence that he is wrong if you will accept as evidence what he won’t. It is like smoking and cancer — I suspect since we have no randomized clinical trials he is still skeptical of that relationship. And probably thinks that global warming is just a normal trend (which of course it could be).

But randomized trials in education are really hard to come by. And a whole set of other studies suggest that important outcomes do accrue to social programs of the kind he does not like.

For example, if Harlem Children’s Zone can reduce mobility (I don’t know if it does, but I suspect that parents will stay put if their kids and their social needs are better met in the zone) we can improve school achievement.

A reduction in family mobility helps the mobile kids achieve more and increases the test scores of the stayers, as well as the movers, probably because their teachers are able to concentrate better on the stuff they want to teach rather than making up work for the new kids all the time. A fact that Whitehurst knows is that after three moves, the odds of graduating high school go way down as does school achievement. Black kids move about twice as often as white kids.

With medical attention also comes greater school attendance and that has known effects on classes for the absentee kid and the ones that show up all the time. Will Whitehurst say attendance is irrelevant? I doubt it!

And at least one Title 1 study shows that death rates of poor kids goes down to the average in programs fully funded, and that might be both an achievement of some note and a social good we might want to take pride in, even if the test scores of these kids are not higher than kids in private nursery schools (whose parents are bound to be wealthier).

Certainly with regard to nutrition he is wrong: Here is my favorite study, because it is a randomized study and tells Whitehurst he is full of it. I wrote this.

Berliner, David C. (2009). Poverty and Potential: Out-of-School Factors and School Success. Boulder and Tempe: Education and the Public Interest Center & Education Policy Research Unit. Retrieved July 20, 2010 from

"The easiest way to demonstrate the effect of poor diets, hunger, food insecurity, and related nutritional problems on student achievement is with a powerful example from the research literature. Between 1969 and 1977, Guatemalan children in four villages participated in a randomized clinical trial of a nutritional supplement.

"Some were given atole, a protein-rich enhanced nutritional supplement, while others were given fresco, a sugar-sweetened beverage. When 1,448 surviving participants from both groups, about 70% of the original sample, were interviewed and assessed cognitively at an average age of 32, it was found that those children exposed to atole between birth and age 24 months scored substantially higher on intellectual tests of reading comprehension and cognitive functioning in adulthood than those not exposed to atole. Most important, the cognitive gains were independent of years of schooling.

"Those who began the atole drinks after birth and consumed atole for a few years showed these effects, while those who began the supplement later did not.

"This corresponds to what developmental psychologists tell us: complex and rapid cognitive development takes place during the first three years of life. The advantage in schooling was greater for girls than for boys, but even for boys atole consumption in their first few years resulted in almost a half-year more school attendance than was true of the control children.

"The conclusion is that proper nutrition early in life gives rise to greater intellectual functioning and higher levels of education later in life. It is also worth reporting that the children born of mothers who took either drink received a calorie supplement during their pregnancy. This resulted in a LBW (low birth weight) rate of 9%, while the rate of LBW children among mothers that did not get the caloric supplement was 19%. As detailed in the earlier discussion of birth weight, LBW is related to nutrition during pregnancy and is an out-of-school factor that generates problems for children and the schools they attend."

Here is another nutrition study I cite in the paper I cited above. This one shows effects and demonstrates, simultaneously, how nuts our country us:

"Some schools have figured out that such nutritional deficits are affecting all-important test scores in this age of NCLB high-stakes accountability. So they provide extra rich foods on test days, essentially calorie-loading students to give them the energy they need to perform well. It works.

"Gains of from 4-7% on tests accrue to the schools that calorie-load their children.

Sadly, even knowing that this strategy works during test week, indicating convincingly that a district’s children have trouble performing academic tasks on their inadequate normal diets, most or all of these school districts nevertheless continue with the less-rich diet throughout the rest of the year. They fail to address what they know to be true given their attempt to raise test scores through calorie-loading: many children are getting diets that minimize their opportunities to learn in school."

And a whole set of other studies suggest that important outcomes do accrue to social programs of the kind Whitehurst does not like.

If attention to mental health problems reduces family violence then kids from such families will have scores that go up, but also, class scores will go up if the number of troubled kids goes down. For example, I wrote:

"It should not be surprising to find that domestic violence impairs the ability of parents to nurture the development of their children. Mothers who are abused may be depressed or preoccupied with the violence. In turn, they often appear to be emotionally withdrawn or irritable. They may communicate feelings of hopelessness. The result of familial violence is too often a parent who is less emotionally available to his or her children, or unable to care for the children’s basic needs.

"Battering fathers have been found to be less affectionate, less available and less rational in dealing with their children. Children from families that suffer from violence, from whatever income group and race, often display social and emotional problems that manifest themselves in the schools they attend. Too often these children show higher rates of aggressive behavior, depression, anxiety, decreased social competence, and diminished academic performance. In a study of low-income pre-school children in Michigan, nearly half (46.7%) had been exposed to at least one incident of violence in the family. In fact, estimates are that between 3 million and 10 million children witness family violence each year. That affects the schools.

"Children exposed to violence were found to suffer symptoms that resemble post-traumatic stress disorder. They showed increased rates of bed-wetting or nightmares, and they were at greater risk than their peers of having allergies, asthma, gastrointestinal problems, headaches and flu.

"Further, there is now ample evidence that stress during childhood because of poverty, family violence, parental depression, rejection by caretakers and so forth has physiological effects, changing the hormonal levels and the architecture of a child’s brain. Loving and secure relationships with caregivers early in life lead to mentally and physiologically healthier children.

There are two factors to consider in these depressing data. One is that such families and children are overrepresented among the poor and in the African American community, increasing the difficulty of the instructional and counseling missions of schools that serve those populations.

"Secondly, the effects these troubled children exert on others in the classroom are strong. For example, within an elementary grade cohort, an increase in the number of children from families known to have a history of domestic violence shows a statistically significant correlation to a decrease in the math and reading tests among those students’ peers, and to an increase in disciplinary infractions and suspensions among the peers as well. These negative effects were primarily driven by troubled boys acting out, but the effects were present across gender, racial lines and income levels. The researchers estimate that adding one more troubled boy peer to a classroom of 20 students reduces the overall test scores of boys by nearly two percentile points. Girls seem to be less affected by the presence of another troubled child.

"Overall, however, when another troubled child of either sex is added to a class, the mean test score of the class drops by about two-thirds of a percentile, and the probability that disciplinary infractions will occur increases by 16%. The data analysis even revealed that when a child shares a classroom with a victim of family violence, she or he is likely to perform less well than a sibling who attends the same school but in a different classroom with peers experiencing less domestic violence. This study provides support for teachers and parents who believe that 'One bad apple can spoil the bunch.'

And these negative adult attitudes, though rooted in reality, make the classroom and social lives of children from abusive households even harder, especially in schools with weak counseling programs or few social workers.

Okay — enough of a rant.

A) He overstates the lack of effects, probably because he only believes in randomized studies ...

B) He actually fails to cite evidence in the nutrition area that does refute his claim.

C) He refuses to cite data that is correlational about health and violence that refutes his position. Correlation doesn’t mean causation but often implies it, like cigarettes and cancer.

D) He seems not to care about other outcomes such as health and family stability independent of student test scores. I'll take a few points less on tests if I got a reduction in family violence, a reduction in crime rate, a lower death age for minorities, etc. ...

Oh by the way, I just noticed that he [Whitehurst] uses KIPP schools among the comparison charter schools that do better than the HCZ kids.

But KIPP usually takes about 100 kids in at fifth grade and graduates 60 of them at ninth grade. I bet Canada or even the village idiot can raise school scores more by dropping out the kids that “don’t fit.”

There was more, but you must have the gist by now.

What do you think?

Follow my blog all day, every day by bookmarking And for admissions advice, college news and links to campus papers, please check out our Higher Education page at Bookmark it!

By Valerie Strauss  | July 22, 2010; 6:30 AM ET
Categories:  David Berliner, Equity, Guest Bloggers, Learning  | Tags:  alexander russo, brookings institution report, brookings report and hcz, david berliner, effects of poverty on student achievement, factors in student achievement, harlem children's zone, jay mathews, student achievement, whitehurst report  
Save & Share:  Send E-mail   Facebook   Twitter   Digg   Yahoo Buzz   StumbleUpon   Technorati   Google Buzz   Previous: Do high standards really help kids?
Next: Wake County diversity battle


yeah that is true, major brands do give out free samples of their popular health products best place to get yours is tell your friends and family too

Posted by: briansusan | July 22, 2010 6:51 AM | Report abuse

I am surprised Obama would support anything that takes a child's socio-economic background into account. I thought with Obama only test scores counted, and if teachers didn't produce high tests - they were to be punished.

Anyway, I don't think Harlem Children's zone should be judged by test scores. It sounds like its on the right track.

Posted by: jlp19 | July 22, 2010 10:38 AM | Report abuse

I am always amazed at willing to accept the evidence of the experts that is totally flawed and I glad Valerie Strauss has actually gone to an expert to point out the flaws.

The Washington Post would be well advised to have Valerie Strauss check with experts and review of any article of Mr. Mathews that involves testing or mathematics.
Posted previously in regard to the Brookings study with slight editing.
Mr. Mathews:
So it is still interesting that an institution as influential as Brookings decided to issue a report that says that the Promise Academy is not doing as well as other charters that don't benefit from the safety net services of the zone, and thus may be a failure.
Apparently Mr. Mathews has no concept about children or testing.

One is not testing the plumbing pipe manufactured by a company.

Testing two populations of children from different schools only indicates their scores on the test.

It can not tell you anything about the effectiveness of the methods of the different schools.

Children are not plumbing pipe and contrary to Mr. Mathews all black students are not the same.

School A had better grades than school B.

No one tested the capabilities and skills of the students of school A when these students originally entered School A.

No one tested the capabilities and skills of the student of school B when they entered School B.

The manufacturer testing plumbing pipe will only use material that are similar. Nothing can by gained from testing when there is guarantee that the materials that are required for manufacturing plumbing pipe are not the same.

For all we know school B did a better job than School A that had higher scores since it might have had students that entered their school with far lower capabilities and skills than the students that entered school B.

Test only tell you how well or poorly someone has done on a test.

This entire concept of testing of students indicating the effectiveness of different schools is flawed.

Without any knowledge of the skills and capabilities when students enter the school you can not use testing to make any statements regarding the school system.

To judge the effectiveness of school systems with tests you need to have for every student the baseline test of their abilities and capabilities when they entered the school system.

Without these baselines test the only thing that can be said when the students of school A do better on a test than the students of school B is that the students of school A did better on a test than the students of school B.

Posted by: bsallamack | July 21, 2010 2:13 PM

Posted by: bsallamack | July 22, 2010 2:43 PM | Report abuse

Thanks for David Berliner's incisive response...
With regard to using KIPP schools in the comparison, I would suggest that KIPP functions somewhat like a social services intervention. By keeping kids at school for nearly 12 hours per day, along with evening phone access to teachers, and Saturday school, isn't that in essence "taking over for the family", or supplanting the family responsibility significantly? Wouldn't that likely benefit students from problematic homes? Though as Berliner suggests, it also seems likely that KIPP lets go of students whose problems are particularly severe and whose parents do not have the capacity to meet the demands of the school.

Posted by: Incidentally | July 22, 2010 2:49 PM | Report abuse

Great reporting once again, Valeria. You knocked the ball right out of the park with this one. I agree bsallamack in respect to Jay Mathews.

Posted by: lacy4 | July 22, 2010 9:56 PM | Report abuse

I believe Dr. Berliner made an error about Kipp schools when he said that only 60 out of a hundred students complete the program. The correct statistic is that KIPP schools typically LOSE 60 percent of their students in that period.

A November, 2008, summary of KIPP by Dr. Jeffrey Henig cited a Washington Post news account that the KIPP Bridge College Preparatory school in Oakland, California, lost 62 of 87 students who enrolled in 2003; the KIPP school studied in Memphis, Tennessee, lost 33 of 49 students from fifth to eighth grade; the study of KIPP schools in San Francisco found that "60% of students who started KIPP in fifth grade had left before the end of eighth grade."

More importantly, both KIPP schools and other programs such as HCZ only enroll and sustain students who are likely to succeed. Students with behavioral and cognitive problems are not likely to enroll in charter schools that enforce strict behavioral contracts.

The real secret to the success of these schools, that has been repeatedly demonstrated by research, is that they enroll students who typically have higher academic abilities than the average school and then lose most of their lower scoring students over time so that they only report their successes and not their failures.

Public schools accept all students and report on all students, including the special education students that rarely enroll in charter alternatives.

Posted by: zoniedude | July 23, 2010 3:16 PM | Report abuse

It is an absolute fallacy that KIPP schools do not take any students unless they are academically gifted. I am very familiar with several KIPP schools and I know for a fact that they accept everyone, including special needs students. In truth, many of the kids who aren't there in 8th or 9th grade have been taken out by parents who don't like the system, for whatever reason. Those spots are filled with additional kids on a continual basis, at all grades.

If you don't believe in KIPP, ask yourself: is it because you don't believe in their structure and system, or because you don't believe in the kids? Because KIPP takes some very average students from very destitute backgrounds, and gives them a shot at a future.

Posted by: samslaw25 | July 26, 2010 6:42 PM | Report abuse

The comments to this entry are closed.

RSS Feed
Subscribe to The Post

© 2010 The Washington Post Company