At RSF, Greg Duncan (University of California, Irvine), a current Visiting Scholar and co-editor of Whither Opportunity?, is studying programs that aim to increase the school success of low-income children and youth. He is examining whether these interventions remain effective in the long run, and whether they help restore some of the social mobility that has been lost over the past several decades. In a new interview with the foundation, Duncan discussed patterns of fadeout and persistence across different educational interventions, and how to sustain gains made by early-childhood interventions.
Q. Your current research at RSF analyzes the impacts of a number of educational interventions for children and adolescents. What do longitudinal data reveal about the longer-term effects of different interventions? Which interventions have succeeded?
Hundreds of educational interventions have been developed and evaluated over the past several decades. The vast majority—whether they address academic skill-building, social or emotional capacity-building, or students’ social-psychological wellbeing—measure their success by looking only at their impacts at the end of the program. But what happens in the years that come after the end of the program? A systematic look at evaluations of early childhood education programs with follow-up assessments published between 1960 and 2007 showed a geometric decline in impacts after the end of the programs—a very disappointing result.
One example is the Building Blocks program. Building Blocks is a very creative early mathematics curriculum in which preschool children received 15-20 minutes of math education every day through different games and puzzles. The curriculum is play-based, so the children in the program have fun as they learn about math. In most of the sites it has been tested, the curriculum produced big impacts at the end of preschool—in one case the program had closed about two-thirds of the school entry math achievement gap between high- and low-income children. But program impacts also faded out rather quickly: By third grade there were no significant differences at all in the math achievement of children who had been in the pre-K Building Blocks treatment and control groups.
The Double Dose algebra program produces very different patterns of impacts. Double Dose consisted of providing two class periods of algebra, rather than just one, for ninth graders in Chicago public schools. Students were selected for the program based on their scores on a state math test at the end of eighth grade—virtually all students scoring below the median of the test-score distribution were enrolled in Double Dose and virtually none of those scoring above it were enrolled. The two classes were usually taught by the same teacher to the same students. The Double Dose curriculum tried to emphasize building a conceptual understanding of algebra, rather than just cranking out the formulas.
In contrast to Building Blocks, the Double Dose intervention led to some longer-term improvements for students. Compared with children whose eighth grade math scores were just above the threshold, students in the Double Dose program kids in the program were more likely to have passed algebra, and also geometry; more of them graduated high school, and more of them went to college.
So why was there fadeout in the case of Building Blocks, yet little evidence of fadeout with Double Dose? That question led us to think more broadly about the circumstances under which interventions might be expected to produce persistent versus ephemeral impacts.
Q. What accounts for the wide range of persistence and fadeout patterns between these programs? Why do some impacts persist while others don’t?
The explanation that occurs most naturally to economists, and perhaps others as well, is that if you can just find the right skill or capacity and boost it early, it will provide a lifetime of benefits for students. “Grit” is often thought to be one example. In the case of math, it is possible that if you start building math skills in pre-K in a way that gets students interested in math and helps them understand the basic concepts of math, they’ll be ahead going into kindergarten and will be able to take better advantage of what kindergarten, first grade, second grade and so on have to offer. Of course, we didn’t see this happen with Building Blocks. That isn’t to say that early interventions are never effective, but it does lead us to ask: what are the characteristics of a skill that will generate persistent academic benefits?
We came up with the idea of “trifecta” skills—skills that combine three key characteristics in a way that is likely to generate longer-term impacts. First, the skill has to be malleable—it has to be something that can change with an intervention or program. Next, it has to be fundamental for student success. And then the kicker is that it also has to be a skill that would not have developed anyway for kids who weren’t in the intervention. Building Blocks didn’t satisfy the third criterion because the children in the program were learning math skills in pre-K that most students would soon learn in kindergarten or first grade. If kindergarten instruction is geared toward advancing the math skills of children coming in with low but not higher levels of math skills, then the children with low initial skills will catch up with the Building Blocks children.
Personality skills—like conscientiousness or grit—and IQ have very strong correlations with school and career success. So it’s likely that if you could somehow boost IQ permanently or make someone truly more conscientious, then would lead more successful lives. Skills like these are therefore fundamental and would probably not have developed without an intervention. Yet, IQ and conscientiousness are not trifecta because they aren’t very malleable—we don’t know how to change them with the kinds of interventions that schools might offer.
So what satisfies all of the trifecta characteristics? It’s not likely to be very many of the early skills that kids acquire, but rather, higher-level skills that are taught later in elementary school. Learning fractions or algebra are examples. Many children never learn fractions, and a pretty shocking percentage of community college students can’t solve simple fraction problems. On the literacy side, writing and other communication skills are increasingly important in today’s workplace, as is a sufficiently large vocabulary that enable students to understand advanced texts. These types of skills are all important for success, are malleable, and most would not develop in the absence of an intervention. On the social-psychological side, there’s been interesting recent intervention research showing that boosting motivation and “mindset” in both secondary and post-secondary school settings can generate benefits that persist at least a few years after the end of the intervention.
Q. How can looking at these programs' fadeout patterns help us better construct and implement effective educational interventions? What kinds of broader education policies could help kick start new interventions that build long-term skills for students and ultimately allow schools to foster social mobility?
Studying fadeout patterns can lead us to spend money on programs more efficiently—it’s important to find ways of insuring that impacts from educational interventions don’t fade out. One way to do that is to think hard about trifecta skills and build those up. Another is to rethink how we make decisions about grade failure and special education, which can trap students in lower-achieving tracks.
A third way is to think about ways to generate school environments that function as “charging stations” that sustain impacts of early programs. In the case of Building Blocks, this amounts to asking how might we structure kindergarten and first grade math instruction in ways that would allow kids who have been in Building Blocks to sustain their gains? This could mean building up Gifted and Talented tracks for kids, or providing more training for teachers on how to better instruct both the advanced and non-advanced kids together in their classrooms, especially for kindergarten. You also want kindergarten and first-grade teachers to know what their students experienced in their pre-K years. In many states, pre-K isn’t part of the education department, so it’s essential to bring pre-K into the K-12 education bureaucracy in as integrated way as possible. Pre-K and early grade curricula needs to be aligned. Teachers need to talk to one another, and even take turns teaching in each others’ classrooms in order to understand the nature of student math and literacy learning trajectories across the early grades. These are the types of changes that could help sustain gains from early childhood interventions.