November 2010 Archives

A new randomized control trial in Mississippi has found that a good kindergarten literacy program can boost disadvantaged students' vocabulary in kindergarten by as much as an extra month of school.

Early childhood programs like Mississippi's have focused heavily on early vocabulary for decades, with growing urgency since a seminal 1995 University of Kansas study showed children of parents on welfare enter school knowing about 525 words, less than half of the 1,100-word vocabulary of children of parents in professional jobs.

The Regional Educational Laboratory Southeast, housed at the SERVE Center of the University of North Carolina at Greensboro, evaluated the Kindergarten PAVEd for Success program, which trains teachers to supplement their normal literacy instruction. Pam Finney, the research management leader for the study, said the program was purposely "not a very complicated intervention," and it helps teachers engage in the same complex conversations that the Kansas study showed professional parents have with their children, "introducing 50 cent words as opposed to 25 cent words," as Ms. Finney put it.

Each teacher gets a list of thematically related and complex words; for example, "temperature," "exhaust," "steam," and "boil," or "pineapple," "banana," and "kiwi." The teacher reads stories that incorporate the words with the students and opens conversations with the students.

"One of the strategies is building bridges, having conversations with students whatever they want to talk about," Ms. Finney explained. "The teacher learns how to have these conversations. Take 'apple,' 'banana' and 'Kiwi.' Students in the Delta may never have heard of a kiwi or seen the fruit. So the teacher shows them and they talk about it."

Researchers tracked nearly 1,300 kindergarteners at 30 Mississippi Delta school districts, in which 128 kindergarten classes were randomly assigned to either use the program or teach literacy as they normally would. Teachers in the program received training but were allowed flexibility to implement it. All of the schools had to have at least 40 percent of their students in poverty, and both groups of children were similar demographically.

The researchers found children who participated in K-PAVE had an expressed vocabulary one month ahead in vocabulary development and academic knowledge at the end of kindergarten compared with students in the control group, as measured by a normed test. The students showed no significant difference in listening comprehension skills.

"These students who were below the norm for vocabulary to start, they're one month closer to the norm, one month closer to those middle-class kids," said Ludy van Broekhuizen; the executive director for SERVE Center and the REL's director. "To actually get an impact on an intervention that required such a small effort on the part of the district is sort of remarkable in some ways."

Teachers trained in the program were significantly more likely than the control-group teachers to include activities focused on students' vocabulary and comprehension development, but they did not show significantly more instructional or emotional support for students.

The researchers have just submitted a follow-up study on the children's literacy skills by the end of 1st grade, but they wouldn't share those details yet. Because the students in the K-PAVE study improved in vocabulary, but not in comprehension, compared to their peers, I'd be interested to see what a follow-up study on these kids would show. Considering kindergarteners and 1st graders are just learning to read, would a one-month edge be enough to boost these students reading development, get them moved to more advanced groups, and so on? It would be interesting to find out. Moreover, since the original "vocabulary gap" study focused on parents' conversations, not teachers', I'd be interested in whether similar training could help parents improve their conversations with their children, too.

Posted by Sarah D. Sparks at 7:00 AM | Permalink | Leave a comment | Recommend

High-achieving schools can exacerbate the social cost to high-achieving students of color for allegedly "acting white" among their peers, according to a study in the November/December issue of Child Development.

Thomas E. Fuller-Rowell, now an Institute for Social Research fellow at the University of Michigan, led researchers at Cornell University in analyzing data on more than 100,000 students of black, white, Asian, Native American and Hispanic students in grades 7-12 from the National Longitudinal Study of Adolescent Health. They compared students' grade point averages with a measure of students' feelings of loneliness, social support, and sense of belonging.

"The social cost of achievement has been a relatively neglected topic in research on the achievement gap," Fuller-Rowell told me. "We already know that social acceptance is one of the primary concerns of adolescence. If achievement comes at a social cost, there are obviously going to be differences in teenagers' motivation to achieve." Academic achievement, he said is "not just correlated with social acceptance, it actually predicts social acceptance in certain school contexts."

The authors found black and Native American adolescents each had significantly higher social costs associated with academic success than did white students, and the social cost was greatest for students who were part of a racial minority in a high-achieving school. Interestingly, this occurred whether or not white students or another racial group made up the majority of the students.

"The main interpretation is these schools are likely to create a more competitive environment, and any competitive environment will increase tensions between groups," Fuller-Rowell said. Moreover, in a highly competitive school, "It's difficult to achieve highly without engaging in behaviors that are visible to peers" such as speaking out in class or participating in clubs, he said.

While Hispanic students as a whole did not take a social hit for high achievement, Mexican-American students followed the same trajectory as black and Native American students of showing greater social alienation in high-achieving schools in which they were the racial minority, but gaining social status with academic achievement in high-achieving, majority Hispanic schools.

By contrast, students attending a high-achieving school in they made up the majority, such as black students at a high-achieving, mostly black school, did not feel stigmatized for excelling.

Posted by Sarah D. Sparks at 8:22 AM | Permalink | Leave a comment | Recommend

Clarification: An earlier version of this post said that Hall announced her resignation; she announced that she will leave at the end of her current contract.

The Senate has been dithering since June over whether or not to confirm Beverly L. Hall's status as a member of the National Board for Education Sciences, the advisory group for the Education Department's research arm, and developments over the weekend may have complicated matters even further.

As the Associated Press reported this morning, Hall announced over the weekend that she would step down as Atlanta's schools superintendent after her current contract expires in June 2011. The announcement comes in the middle of a sweeping state investigation of alleged cheating during the 2009 Criterion-Referenced Competency Test, the state's achievement test.

The American Association of School Administrators had named Hall the 2009 Superintendent of the Year for her work in the district, which they called a model for urban school reform. Hall introduced many programs similar to the reforms seen in New York's Children First initiative, San Diego's Blueprint for Student Success, and elsewhere, including: Project GRAD's intensive reading blocks; the Bill and Melinda Gates Foundation-sponsored reorganization of large comprehensive high schools into small academies; and General Electric Foundation-sponsored science and math programs. Hall was intended to provide a practitioner's viewpoint on the board, which recently has grappled with questions of how researchers should work with districts to evaluate reforms.

Institute for Education Sciences Director John Q. Easton said that it is too early to know if this will have any any effect on her nomination and much may depend on what she ends up doing next. We'll have to wait and see.

Posted by Sarah D. Sparks at 9:43 AM | Permalink | Leave a comment | Recommend

Researchers and policymakers have urged parents for years to read to their young children, even infants, to help them develop better vocabulary and reading readiness. Now, a new study by the University of Chicago suggests parents should be talking to their toddlers about numbers, too.

The study, "What Counts in the Development of Young Children's Number Knowledge?," in the current issue of Developmental Psychology, suggests there are big differences in the amount of number-related words parents use in regular conversation with their children, and this can have a big effect on a child's numeracy, even before formal number instruction in preschool.

Susan C. Levine, a psychologist at the University of Chicago, and a team of researchers found toddlers whose parents talked with them frequently about numbers were better able to understand one of the foundation principles of early math: the cardinal number principle, i.e., the understanding that the number "six" represents a set of six items. According to researchers, children learn the abstract meaning of a given number separately from simply learning to count to that number.

The team studied 44 preschool children interacting with their parents during everyday activities in five 90-minute taped home visits conducted every four months from the time the children were 14 to 30 months old. The researchers then coded the number of times a parent used a number-related word, such as pointing to a series of toys on the floor and saying, "There are four trucks."

Researchers found parents wildly varied in the amount of number-related words they used around their children, from as few as four to as many as 257 &mdash which would translate to a range of 28 to 1,799 number-related words used per week between the most and least vocal parents. Moreover, Ms. Levine's team found children whose parents used more number words in discussions when the child was 14 to 30 months old were more likely at 46 months old, or just at preschool age, to be able to answer accurately when shown two sets of four and five blocks and asked to point out the set of five.

"By the time children enter preschool, there are marked individual differences in their mathematical knowledge, as shown by their performance on standardized tests," Ms. Levine said in a statement on the study. "These findings suggest that encouraging parents to talk about numbers with their children, and providing them with effective ways to do so, may positively impact children's school achievement."

Posted by Sarah D. Sparks at 2:38 PM | Permalink | Leave a comment | Recommend

New York City's massive Children First initiatives under Mayor Michael Bloomberg and outgoing schools chancellor Joel I. Klein already have a head start on being among the most well-documented urban system reforms of the No Child Left Behind era, as a series of new studies highlight.

Yet while attending the research preview in New York on Wednesday, I was struck by how well the city has set itself up for cycles of reforms, research, and renewal that could become the model for school systems nationwide, regardless of what people think of the specific initiatives set up by Klein and Company.

"You cannot pull off education reform without being comprehensive and without being in a constant cycle of improvement," said James S. Liebman, a Columbia Law School professor and the designer of New York City's accountability system. "So you're going to put research at a disadvantage from the beginning because it has to keep changing. ... You contract with researchers to study something and by the time they finish studying it, it's moot because everything has changed."

Mr. Liebman said he is working on a model for future evaluations based on his experience in New York. The goal is to help other district officials and researchers structure both reform initiatives and evaluation studies which take into account that changing environment. Hopefully, districts will end up with more accurate insights into what works and doesn't work in comprehensive reform packages like New York's.

Children First was and remains a giant interconnected web of federal, state and local policies, labor agreements, new curricula and materials, and personnel training and resources all playing off one another, which would normally make it next to impossible for researchers to evaluate individual aspects. Yet the city's education department rolled out several initiatives, such as teacher pay and evaluation data reports and student incentive programs, as randomized pilot programs before expanding them districtwide, making the initiatives' effects easier to study.

"The DOE had enough forethought and was brave enough to do these prospective studies," noted Jonah Rockoff, an education economist at Columbia Business School in New York. As a result, he said, "the DOE has become a bit of a laboratory for people across the country to look at what works."

Posted by Sarah D. Sparks at 3:45 PM | Permalink | Leave a comment | Recommend

If there's one complaint in the education research community that never loses steam (well, besides the call for more funding) it's the frustration that great research rarely seems to translate into great policy or programs. In 2007, the National Research Council of the National Academies of Science launched a Standing Committee on Social Science Evidence for Use to try to understand why researchers have difficulty communicating their findings to policymakers. The committee is nearing the release of its final report, and I, along with the Academy of Education's new post-doctoral fellows, last week got a sneak preview of what's in the report.

Committee member Kenneth Prewitt, a public affairs professor at Columbia University in New York City, said academic researchers often criticize think-tank and other policy researchers without considering what can be learned from their approach. "We have a public policy industry: 300 think tanks in this town (Washington D.C.) alone, public policy schools, university policy centers.... We have an industry that wants to make public policy in this country; why then do we say we make social science and it's not being used?" Prewett said during a discussion on the committee's study. "If [policymakers] want social science and they're paying for it and we think it's not being used, we need to know why. ...If we start from what is being used, that will give us insight into [it]."

The committee is exploring how research evidence interacts with politics, policy processes, and other systems. Miron L. Straf, the deputy director of the National Academy of Sciences' division of behavioral and social sciences and education suggested researchers and policymakers talk past each other in part because they often look for different things from ostensibly the same study.

"Establishing cause and effect is very different from determining what works," Mr. Straf said. "We need theories of action that explain the causal mechanism behind the implementation and we need theories of implementation."

Mr. Straf illustrated his argument with the following very concise and helpful comparison table:

The final report is due out in the next few months.

Posted by Sarah D. Sparks at 1:01 PM | Permalink | Leave a comment | Recommend

Can a motley group of teachers, architects, concerned parents, and folks who wandered in off the street build a better classroom?

I joined an auditorium full of education aficionados at Slate's 2010 Hive Project on education in Washington, D.C., on Monday to see what would shake out.

Crowdsourcing—the notion that problems may be solved better and more rapidly by groups rather than individuals—has long been the darling of the Web 2.0 crowd. A natural fit with theories of group learning, it has been making inroads in the traditionally siloed education field through projects like Duke University's peer-graded Internet course and Curriki's open-source lesson-planning and textbook community.

Slate's editors sent out an open request for readers to submit ideas for a 5th-grade classroom of the future, no budget or personnel constraints required. Before the live event Monday night, editors had received more than 350 ideas, which they and readers narrowed down to 10. Those entries seemed interesting, but most weren't really unheard of: Outdoor classrooms, multipurpose spaces, Montessori-style group learning will sound familiar to most educators and researchers. One really radical idea, which called for hooking students into sensors to monitor their brain waves, would make a great summer blockbuster but probably wouldn't pass muster at your average parent-teacher organization or school board budget meeting.

So there we were, with an expert panel including Jim Shelton, the Education Department's innovation guru; Justin Cohen, president of the Mass Insight turnaround group; David Ethan Greenberg, board chairman of the Denver School of Science & Technology, and a slew of others. Greenberg and several other panelists argued that building standards virtually ensure classes will be outdated almost as soon as they open: "The mindset in districts is they want you to build to hundred-year building standards," Greenberg said, "and we said we have no idea what education is going to look like 15 years from now; why would we invest in these double-wide corridors that come from the factory model ....?"

The most interesting conversation for me was the small group discussion on classroom changes under $1,000.

We had a couple of teachers, the head of a community tutoring group, several businesspeople and a handful of others. A 9th-grade English teacher said she gets $175 a year for class supplies; a Prince George's County, Md., kindergarten teacher said he gets none. Both already try to create cobbled-together versions of the high-concept classrooms of the future: She arranges classroom areas for different projects; he plugs his laptop from home into a projector to show his students online resources. Both begged for basic improvements already known in the education field, like data profiles of incoming students or desks uncoupled from chairs to ease moving for group work. (Slate Editor-in-Chief David Plotz, who sat in on the group, looked shocked that classes still used such desks.) One tourist attending the forum on a whim suggested exchanging the desks for lap-tables and using chairs on rollers.

Plotz, who has held a similar crowdsourcing event on energy policy, told me half-jokingly afterward that he always hopes at these events to have someone jump up and announce that one brilliant idea that will solve everything wrong in education. Yet maybe the strength of crowdsourcing, if there is one, comes not from a single silver-bullet but an understanding of all the millions of little irritations teachers and students face every day and the ways to make classes a little more livable. I'll be interested to see what ideas come out of the project going forward.

Posted by Sarah D. Sparks at 7:47 AM | Permalink | Leave a comment | Recommend

It looks like there will be some continuity for the National Board for Education Sciences, which advises the Education Department's research agency, as it awaits new members' Senate confirmations.

Jon Baron, the current board vice chairman and one of the few NBES members who has been involved since the board's founding in 2004, will step in as chairman from current head Eric A. Hanushek, a Stanford University economist, whose term expires on Nov. 28. New member Bridget T. Long, an education and economics professor at the Harvard University Graduate School of Education, will take over as vice chairman.

Baron is president of the Washington-based Coalition for Evidence-Based Policy, which under his tenure worked with the federal Office of Management and Budget to develop the Program Assessment Rating Tool, which has been used to gauge the effectiveness of federal programs for budget purposes. He also serves on the National Academies' Committee on Capitalizing on Science, Technology, and Innovation, and is an honorary fellow of the Academy of Experimental Criminology. On the education sciences board, Baron has championed the use of experimental design in education studies.

"When practitioners, policymakers, educators think of a randomized controlled trial, they often think of a big expensive enterprise that will involve 100 different schools and measure a lot of different outcomes and will produce a 300-page report and cost a lot of money," he said at the board's last meeting. "Good randomized evaluations don't have to be that way. The biggest cost in the evaluations is tracking students over time, but in many cases now that can be done using administrative data like student test scores. That basically takes the biggest cost in these trials and makes it a nominal cost."

Baron advocates helping district leaders conduct their own small-scale research to answer their own policy questions. He pointed to a superintendent in Seminole County, Fla., who wanted to choose a remedial reading program for 9th graders. The superintendent took 1,500 of his incoming freshman who scored poorly on the state assessments in the previous year and randomly assigned them to one of the three programs, then looked at the students' test scores at teh end of that year and the following year. "The total cost of that evaluation was about $50,000," Baron said.

Posted by Sarah D. Sparks at 8:23 AM | Permalink | Leave a comment | Recommend

The National Board for Education Sciences this afternoon unanimously approved new research priorities for the U.S. Department of Education's research arm that are intended to make education studies more relevant to educators and help practitioners become more involved in developing and using research.

The Institute of Education Sciences' topics of study won't change much under the new priorities; they include educational processes, instructional innovations and teacher recruiting, retention, training and effectiveness (adopting the stimulus bill's focus on teacher effectiveness over the older "teacher quality.") Yet the new priorities put greater emphasis on putting their research findings into context, "to identify education policies, programs, and practices that improve education outcomes, and to determine how, why, for whom, and under what conditions they are effective."

IES has set as a priority identifying new and rigorous methods to measure outcomes in education research—another nod to critics who call for the institute to broaden the scope of what it considered the highest quality research beyond randomized controlled trials. Yet IES Director John Q. Easton tweaked the final proposed priorities for in response to board members' concern that an earlier draft appeared to move away from IES's focus on rigorous experimental design studies. The final approved document includes an early confirmation that, "The Institute seeks to understand causal linkages to the greatest extent possible by conducting or sponsoring rigorous studies that support such inferences."

In light of IES' decision to accept some additional research methods as meeting the highest quality bar, board member Adam Gamoran said it is particularly important that the researchers "only subscribe causal inferences to studies with methodologies that allow it. This language addresses my concerns."

The institute will also focus on building partnerships with educators and the community to develop more "analytic capacity" at the local level—something that Mr. Easton said will be part of the next iteration of regional education laboratories, as well. Board member and superintendent Margaret R. McCleod said she "particularly appreciate[s] the fact that [IES] included in the stakeholders parents and students themselves."

The new priorities will be used to craft requests for proposals for new grant competitions in January, and Mr. Easton told the board that he will link descriptions for new grant competitions back to them.

The final priorities are not yet online but are expected to be posted here.

Posted by Sarah D. Sparks at 12:39 PM | Permalink | Leave a comment | Recommend