Search EdWorkingPapers
Search EdWorkingPapers by author, title, or keywords.
Search
Methodology, measurement and data
Community schools are an increasingly popular strategy used to improve the performance of students whose learning may be disrupted by non-academic challenges related to poverty. Community schools partner with community based organizations (CBOs) to provide integrated supports such as health and social services, family education, and extended learning opportunities. With over 300 community schools, the New York City Community Schools Initiative (NYC-CS) is the largest of these programs in the country. Using a novel method that combines multiple rating regression discontinuity design (MRRDD) with machine learning (ML) techniques, we estimate the causal effect of NYC-CS on elementary and middle school student attendance and academic achievement. We find an immediate reduction in chronic absenteeism of 5.6 percentage points, which persists over the following three years. We also find large improvements in math and ELA test scores – an increase of 0.26 and 0.16 standard deviations by the third year after implementation – although these effects took longer to manifest than the effects on attendance. Our findings suggest that improved attendance is a leading indicator of success of this model and may be followed by longer-run improvements in academic achievement, which has important implications for how community school programs should be evaluated.
Increasing numbers of students require internet access to pursue their undergraduate degrees, yet broadband access remains inequitable across student populations. Furthermore, surveys that currently show differences in access by student demographics or location typically do so at high levels of aggregation, thereby obscuring important variation between subpopulations within larger groups. Through the dual lenses of quantitative intersectionality and critical race spatial analysis, we use Bayesian multilevel regression and census microdata to model variation in broadband access among undergraduate populations at deeper interactions of identity. We find substantive heterogeneity in student broadband access by gender, race, and place, including between typically aggregated subpopulations. Our findings speak to inequities in students’ geographies of opportunity and suggest a range of policy prescriptions at both the institutional and federal level.
How much does family demand matter for child learning in settings of extreme poverty? In rural Gambia, families with high aspirations for their children’s future education and career, measured before children start school, go on to invest substantially more than other families in the early years of their children’s education. Despite this, essentially no children are literate or numerate three years later. When villages receive a highly-impactful, teacher-focused supply-side intervention, however, children of these families are 25 percent more likely to achieve literacy and numeracy than other children in the same village. Furthermore, improved supply enables these children to acquire other higher-level skills necessary for later learning and child development. We also document patterns of substitutability and complementarity between demand and supply in generating learning at varying levels of skill difficulty. Our analysis shows that greater demand can map onto developmentally meaningful learning differences in such settings, but only with adequate complementary inputs on the supply side.
Data science applications are increasingly entwined in students’ educational experiences. One prominent application of data science in education is to predict students’ risk of failing a course in or dropping out from college. There is growing interest among higher education researchers and administrators in whether learning management system (LMS) data, which capture very detailed information on students’ engagement in and performance on course activities, can improve model performance. We systematically evaluate whether incorporating LMS data into course performance prediction models improves model performance. We conduct this analysis within an entire state community college system. Among students with prior academic history in college, administrative data-only models substantially outperform LMS data-only models and are quite accurate at predicting whether students will struggle in a course. Among first-time students, LMS data-only models outperform administrative data-only models. We achieve the highest performance for first-time students with models that include data from both sources. We also show that models achieve similar performance with a small and judiciously selected set of predictors; models trained on system-wide data achieve similar performance as models trained on individual courses.
A student's class rank has important short and long-term effects on important educational outcomes. Despite our growing understanding of these rank effects, we still do not know how early in a child's academic career they begin. To address this, I use data from the Tennessee STAR project, which randomly assigned over 6,323 kindergarteners to classroom environments, to study the impact of kindergarten class rank on a host of short and long-run outcomes. I find a strong, causal relationship between one's kindergarten classroom rank and subsequent test scores, high school achievement and performance on college entrance exams. I also find that having a higher rank in kindergarten causes an increase in study effort, value of school and initiative in the classroom. I also leverage the design of project STAR to test various mechanisms and address several outstanding issues in the rank literature, including the role of tracking, parental effort and teacher-level characteristics in driving the effects of class rank.
Career and technical education (CTE) has existed in the United States for over a century, and only in recent years have there been opportunities to assess the causal impact of participating in these programs while in high school. To date, no work has assessed whether the relative costs of these programs meet or exceed the benefits as described in recent evaluations. In this paper, we use available cost data to compare average costs per pupil in standalone high school CTE programs in Connecticut and Massachusetts to the most likely counterfactual schools. Under a variety of conservative assumptions about the monetary value of known educational and social benefits, we find that programs in Massachusetts offer clear positive returns on investment, whereas programs in Connecticut offer smaller, though mostly non-negative expected returns. We also consider the potential cost effectiveness of CTE programs offered in other contexts to address questions of generalizability.
Schools often track students to classes based on ability. Proponents of tracking argue it is a low-cost tool to improve learning since instruction is more effective when students are more homogeneous, while opponents argue it exacerbates initial differences in opportunities without strong evidence of efficacy. In fact, little is known about the pervasiveness or determinants of ability tracking in the US. To fill this gap, we use detailed administrative data from Texas to estimate the extent of tracking within schools for grades 4 through 8 over the years 2011-2019. We find substantial tracking; tracking within schools overwhelms any sorting by ability that takes place across schools. The most important determinant of tracking is heterogeneity in student ability, and schools operationalize tracking through the classification of students into categories such as gifted and disabled and curricular differentiation. When we examine how tracking changes in response to educational policies, we see that schools decrease tracking in response to accountability pressures. Finally, when we explore how exposure to tracking correlates with student mobility in the achievement distribution, we find positive effects on high-achieving students with no negative effects on low-achieving students, suggesting that tracking may increase inequality by raising the ceiling.
Using daily lunch transaction data from NYC public schools, I determine which students frequently stand next to one another in the lunch line. I use this `revealed' friendship network to estimate academic peer effects in elementary school classrooms, improving on previous work by defining not only where social connections exist, but the relative strength of these connections. Equally weighting all peers in a reference group assumes that all peers are equally important and may bias estimates by underweighting important peers and overweighting unimportant peers. I find that students who eat together are important influencers of one another's academic performance, with stronger effects in math than in reading. Further exploration of the mechanisms supports my claim that these are friendship networks. I also compare the influence of friends from different periods in the school year and find that connections occurring around standardized testing dates are most influential on test scores.
Reclassification can be an important juncture in the academic experience of English Learners (ELs). Literature has explored the potential for reclassification to influence academic outcomes like achievement, yet its impact on social-emotional learning (SEL) skills, which are as malleable and important to long-term success, remains unclear. Using a regression discontinuity design, we examine the causal effect of reclassification on SEL skills (self-efficacy, growth mindset, self-management, and social awareness) among 4th to 8th graders. In the districts studied, reclassification improved academic self-efficacy by 0.2 standard deviations for students near the threshold. Results are robust to alternative specifications and analyses. Given this evidence, we discuss ways districts might establish practices that instill more positive academic beliefs among ELs.
Responsive teaching is a highly effective strategy that promotes student learning. In math classrooms, teachers might funnel students towards a normative answer or focus students to reflect on their own thinking, deepening their understanding of math concepts. When teachers focus, they treat students’ contributions as resources for collective sensemaking, and thereby significantly improve students’ achievement and confidence in mathematics. We propose the task of computationally detecting funneling and focusing questions in classroom discourse. We do so by creating and releasing an annotated dataset of 2,348 teacher utterances labeled for funneling and focusing questions, or neither. We introduce supervised and unsupervised approaches to differentiating these questions. Our best model, a supervised RoBERTa model fine-tuned on our dataset, has a strong linear correlation of .76 with human expert labels and with positive educational outcomes, including math instruction quality and student achievement, showing the model’s potential for use in automated teacher feedback tools. Our unsupervised measures show significant but weaker correlations with human labels and outcomes, and they highlight interesting linguistic patterns of funneling and focusing questions. The high performance of the supervised measure indicates its promise for supporting teachers in their instruction.