Social And Emotional Skills In School: Pivoting From Accountability To Development
Our guest authors today are David Blazar and Matthew A. Kraft. Blazar is a Lecturer on Education and Postdoctoral Research Fellow at Harvard Graduate School of Education and Kraft is an Assistant Professor of Education and Economics at Brown University.
With the passage of the Every Student Succeeds Act (ESSA) in December 2015, Congress required that states select a nonacademic indicator with which to assess students’ success in school and, in turn, hold schools accountable. We believe that broadening what it means to be a successful student and school is good policy. Students learn and grow in multifaceted ways, only some of which are captured by standardized achievement tests. Measures such as students’ effort, initiative, and behavior also are key indicators for their long-term success (see here). Thus, by gathering data on students’ progress on a range of measures, both academic and what we refer to as “social and emotional” development, teachers and school leaders may be better equipped to help students improve in these areas.
In the months following the passage of ESSA, questions about use of social and emotional skills in accountability systems have dominated the debate. What measures should districts use? Is it appropriate to use these measures in high-stakes setting if they are susceptible to potential biases and can be easily coached or manipulated? Many others have written about this important topic before us (see, for example, here, here, here, and here). Like some of them, we agree that including measures of students’ social and emotional development in accountability systems, even with very small associated weights, could serve as a strong signal that schools and educators should value and attend to developing these skills in the classroom. We also recognize concerns about the use of measures that really were developed for research purposes rather than large-scale high-stakes testing with repeated administrations.
Ultimately, though, we argue that this narrow focus on accountability misses the forest for the trees and distracts from more important questions of how to effectively support teachers and schools to develop students’ social and emotional skills. As new teachers in New York City and Berkeley, CA, we saw firsthand how students’ own motivation, self-control, and perseverance often were the driving factors behind their success. Consequently, we spent much less of our time focused on improving test scores than we did building positive interactions with students, responding to students’ (mis)behavior, ensuring that they were engaged in the classroom environment, and helping them see how school related to their lives and future aspirations. Even as the culture of standardized testing has grown in U.S. schools over the last several decades, many have urged teachers away from test-centered instruction, arguing that this narrow approach leaves out opportunities for engaging students’ critical thinking skills and building their social and emotional development (see here).
Fortunately, a growing body of research is starting to provide convincing evidence that teachers can have large effects on students’ social and emotional development in addition to their academic performance. We contribute to this evidence and extend it by investigating specific teaching practices that promote these skills. Advancing our understanding of how teachers impact student outcomes beyond test scores and what we can do to support teachers in this work is especially critical as states, districts, and schools simultaneously wrestle with new ways to evaluate their teacher workforces and adopt new content standards that require teachers to incorporate higher-order thinking and social and emotional learning into the curriculum.
Emerging Evidence on Teachers’ Effect on Students’ Social and Emotional Development
Although research traditionally has examined how teachers contribute to students’ performance on standardized tests, several recent studies have begun to document meaningful variation among teachers in their ability to improve a range of other student outcomes (see here, here, and here). Borrowing from the literature on teachers’ “value-add” to student test scores, these studies aim to isolate the unique contribution of teachers to student outcomes in a way that is not confounded with the non-random sorting of teachers to students, the specific set of students in the classroom, or factors beyond teachers’ control. These techniques demonstrate substantive teacher effects on students’ social and emotional development, as well as observed school behaviors including absences and suspensions.
In recent work (see here), we extend the research base with three key findings. Like others who utilize value-added models, we found that teachers have substantial impacts on students’ self-reported self-efficacy, behavior, and happiness in the range of 0.14 to 0.23 standard deviations. This means, that, relative to an average teacher, teachers one standard deviation above the mean in effectiveness (i.e., at the 84th percentile of the distribution of effectiveness) move the medium student up to roughly the 55th percentile of students’ self-efficacy and behavior, and roughly the 60th percentile of students’ happiness. These effects are quite large. Similarly sized teacher effects on students’ test scores (see here) have been used to justify policies to evaluate, compensate, and promote teachers based on these scores.
Our findings also help address a key challenge in the teacher effectiveness literature: Despite efforts to identify characteristics that differentiate between effective and ineffective teachers, the nature of high-quality teaching (i.e., what teachers do to improve student outcomes) largely remains a black box. Indeed, in our analyses, observations of teachers’ classrooms identified only one element – the extent to which they made errors in their instruction – that predicted gains in students’ math performance; this relationship was small relative to the effects described above. Comparatively, we found that several teaching skills were strong predictors of improvements in students’ social and emotional outcomes. Teachers’ emotional support for and interactions with students was related both to their self-efficacy and happiness. Teachers’ classroom organization predicted students’ reports of their own behavior in class. Errors in teachers’ presentation of math content was negatively related to students’ self-efficacy and happiness. These intuitive relationships identify important mechanisms by which teachers impact students’ social and emotional development and, in turn, highlight several skills that could be targeted in both pre-service and in-service training programs.
Finally, we found that teachers were not equally effective at improving all outcomes. Of teachers in the top quintile of effectiveness based on their contributions to students’ mathematics test scores, only 41% were ranked in the top quintile of teachers who developed students’ self-efficacy; 32% were in the lowest two quintiles of effectiveness. We illustrate this weak relationship in Figure 1, which shows the scatterplot and best-fit line between teacher effects on these two outcomes. These findings, along with similar results from studies using additional social and emotional measures (see here), add important empirical evidence on the multidimensional nature of teaching and, thus, the need for researchers, policymakers, and practitioners to identify strategies for improving these skills.
Finding Ways to Support Classrooms Focused on Building Positive Social and Emotional Skills
How, then, can districts and policymakers help schools and teachers support students’ social and emotional development in addition to their academic performance?
The first important step needs to be for states, districts, and schools to articulate clearly what skills they want children to have when they leave school. The Common Core State Standards are one example of educators conveying the central importance of students’ critical thinking and reasoning skills in addition to core content knowledge. We imagine that states will continue to engage in this process as part of ESSA’s requirements. In these and other settings, we encourage stakeholders to think broadly. We know it would be easy to focus on a single or narrow set of social and emotional measures. "Grit" and a "growth mindset" are two examples that have received considerable attention in recent public dialogue (see here, here, here, and here). However, such an approach can act to oversimplify the complex nature of learning. We see this potential narrowing as a real possibility if school accountability systems are the primary mechanisms through which states and districts attempt to promote students' social and emotional development.
Even as we continue to wrestle with important question about which skills to prioritize and measure, there still are many ways we can learn from the data and resources that are currently collected by education agencies. One way would be to leverage administrative data within many agencies to create, as Ryan Knight coined it, a “GRAD” score (grades, retention, attendance, discipline) (see here). Another way would be to draw on data from teacher observations, which have been adopted by many states, districts, and schools as part of new evaluation systems. The evaluative focus of these data mean that teachers often receive a single score on an observation rubric (see here), which may obscure the relationships between specific teaching practices and students’ social and emotional development (see here). However, disaggregated data from these observations often include rich measures of teachers’ skills in discrete areas, much like those captured on the instruments from our research study. Results from our work suggest that these data could be used as proxies for teachers’ ability to improve students’ social and emotional outcomes.
Finally, in order to fully leverage teachers as a resource to improve students’ social and emotional development, we see a pressing need to hold a mirror up to the approaches typically used to support teachers’ work in the classroom. This includes pre- and in-service development programs, as well as curricular materials. Research based predominantly in the human development literature identify several promising school- and teacher-based programs to improve students’ skills in areas beyond their core academic knowledge (see here). However, by and large, the evidence that has emerged has focused on one-time lessons or quick computer programs, rather than efforts to dramatically change teachers’ pedagogy. We see the real potential for teaching social and emotional development as infused into the daily interactions between students, teachers, and content. Social and emotional measures could be used to help teachers diagnose and address the individual factors that may hold students back from fully engaging in school. We know that this is difficult and demanding work that will require coordinating efforts between schools of education, curriculum developers and textbook publishers, and school leaders. Without real support for teachers from all of these sources, we are concerned that social and emotional development might consist of little more than teaching platitudes, with little hope that it might result in actual changes in student outcomes.
One potentially promising approach to developing teachers’ ability to promote social and emotional development is to tap into the existing knowledge of those teachers who have demonstrated the ability to do this work. Here again, measurement of students' social and emotional skills can play a key role outside accountability. Districts and schools could identify “expert teachers” who consistently are able to motivate their students and promote students’ social and emotional development. Such an approach could build upon the growing corps of teacher coaching programs. Leveraging the existing expertise of teachers within schools is another promising and cost-effective approach, given existing resource constraints. For example, peer learning programs could pair a teacher who is skilled in one area but weak in another with a second teacher whose strengths are reversed. Existing experimental evidence for both types of approaches is promising (see here, here, and here).
The complex and multidimensional nature of teaching and of student learning means that improvements in these areas will not be simple or easy. Teaching social and emotional skills also is not a solution to, or a reason to ignore, the challenges many students face outside of school, such as extreme poverty or exposure to violence. However, the alternative is to continue the status quo where some teachers independently choose to integrate social and emotional development into their classrooms and do it well, others do it poorly, and others do nothing at all. No one benefits from this arbitrary approach. Its time we focused on supporting teachers and schools to do this important work.