Search EdWorkingPapers
Search EdWorkingPapers by author, title, or keywords.
Search
EdWorkingPapers
Predictive analytics are increasingly pervasive in higher education. However, algorithmic bias has the potential to reinforce racial inequities in postsecondary success. We provide a comprehensive and translational investigation of algorithmic bias in two separate prediction models -- one predicting course completion, the second predicting degree completion. Our results show that algorithmic bias in both models could result in at-risk Black students receiving fewer success resources than White students at comparatively lower-risk of failure. We also find the magnitude of algorithmic bias to vary within the distribution of predicted success. With the degree completion model, the amount of bias is nearly four times higher when we define at-risk using the bottom decile than when we focus on students in the bottom half of predicted scores. Between the two models, the magnitude and pattern of bias and the efficacy of basic bias mitigation strategies differ meaningfully, emphasizing the contextual nature of algorithmic bias and attempts to mitigate it. Our results moreover suggest that algorithmic bias is due in part to currently-available administrative data being less useful at predicting Black student success compared with White student success, particularly for new students; this suggests that additional data collection efforts have the potential to mitigate bias.
Many dimensions of teacher working conditions influence both teacher and student outcomes; yet, analyses of schools’ overall working conditions are challenged by high correlations among the dimensions. Our study overcame this challenge by applying latent profile analysis of Virginia teachers’ perceptions of school leadership, instructional agency, professional growth opportunities, rigorous instruction, managing student behavior, family engagement, physical environment, and safety. We identified four classes of schools: Supportive (61%), Unsupportive (7%), Unstructured (22%), and Structured (11%). The patterns of these classes suggest schools may face tradeoffs between factors such as more teacher autonomy for less instructional rigor or discipline. Teacher satisfaction and their stated retention intentions were correlated with their school’s working conditions classes, and school contextual factors predicted class membership. By identifying formerly unseen profiles of teacher working conditions and considering the implications of being a teacher in each, decisionmakers can provide schools with targeted supports and investments.
While policymakers have demonstrated considerable enthusiasm for “science of reading” initiatives, the evidence on the impact of related reforms when implemented at scale is limited. In this pre-registered, quasi-experimental study, we examine California’s recent initiative to improve early literacy across the state’s lowest-performing elementary schools. The Early Literacy Support Block Grant (ELSBG) provided teacher professional development grounded in the science of reading as well as aligned supports (e.g., assessments and interventions), new funding (about $1000 per student), spending flexibility within specified guidelines, and expert facilitation and oversight of school-based planning. We find that ELSBG generated significant (and cost-effective) improvements in ELA achievement in its first two years of implementation (0.14 SD) as well as smaller, spillover improvements in math achievement.
Using data from nearly 1.2 million Black SAT takers, we estimate the impacts of initially enrolling in an Historically Black College and University (HBCU) on educational, economic, and financial outcomes. We control for the college application portfolio and compare students with similar portfolios and levels of interest in HBCUs and non-HBCUs who ultimately make divergent enrollment decisions - often enrolling in a four-year HBCU in lieu of a two-year college or no college. We find that students initially enrolling in HBCUs are 14.6 percentage points more likely to earn a BA degree and have 5 percent higher household income around age 30 than those who do not enroll in an HBCU. Initially enrolling in an HBCU also leads to $12,000 more in outstanding student loans around age 30. We find that some of these results are driven by an increased likelihood of completing a degree from relatively broad-access HBCUs and also relatively high-earning majors (e.g., STEM). We also explore new outcomes, such as credit scores, mortgages, bankruptcy, and neighborhood characteristics around age 30.
We examine access to high school Ethnic Studies in California, a new graduation requirement beginning in 2029-30. Data from the California Department of Education and the University of California Office of the President indicate that roughly 50 percent of public high school students in 2020-21 attend a school that offers Ethnic Studies or a related course, but as of 2018-19, only 0.2 percent of students were enrolled in such a course. Achieving parity with economics, a current graduation requirement, requires more than doubling the number of Ethnic Studies teachers relative to 2018-19. We also examine school and community factors that predict offering Ethnic Studies and provide descriptive information about the Ethnic Studies teaching force across the state.
High-quality preschool programs are heralded as effective policy solutions to promote low-income children’s development and life-long wellbeing. Yet evaluations of recent preschool programs produce puzzling findings, including negative impacts, and divergent, weaker results than demonstration programs implemented in the 1960s and 70s. We provide potential explanations for why modern preschool programs have become less effective, focusing on changes in instructional practices and counterfactual conditions. We also address popular theories that likely do not explain weakening program effectiveness, such as lower preschool quality and low-quality subsequent environments. The field must take seriously the smaller positive, null, and negative impacts from modern programs and strive to understand why effects differ and how to improve program effectiveness through rigorous, longitudinal research.
COVID-19 upended schooling across the United States, but with what consequences for the state-level institutions that drive most education policy? This paper reports findings on two related research questions. First, what were the most important ways state government education policymakers changed schools and schooling from the moment they began to reckon with the seriousness of COVID-19 through the first full academic year of the pandemic? Second, how deep did those changes go – are there indications the pandemic triggered efforts to make lasting changes in states’ education policymaking institutions? Using multiple-methods research focused on Colorado, Florida, Louisiana, Michigan, and Oregon, we documented policies enacted during the period from March 2020 through June 2021 across states and across sectors (traditional and choice) in three COVID-19-related education policy domains: school closings and reopenings, budgeting and resource allocation, and assessment and accountability systems. We found that states quickly enacted radical changes to policies that had taken generations to develop. They mandated sweeping school closures in Spring 2020, and then a diverse array of school reopening policies in the 2020/2021 school year. States temporarily modified their attendance-based funding systems and allocated massive federal COVID-19 relief funds. Finally, states suspended annual student testing, modified the wide array of accountability policies and programs linked to the results of those tests, and adapted to new assessment methods. These crisis-driven policy changes deeply disrupted long-established patterns and practices in education. Despite this, we found that state education governance systems remained resilient, and that at least during the first 16 months of the pandemic, stakeholders showed little interest in using the crisis to trigger more lasting institutional change. We hope these findings enable state policymakers to better prepare for future crises.
School principals are viewed as critical actors to improve student outcomes, but there remain important methodological questions about how to measure principals’ effects. We propose a framework for measuring principals’ contributions to student outcomes and apply it empirically using data from Tennessee, New York City, and Oregon. As commonly implemented, value-added models misattribute to principals changes in student performance caused by unobserved time-varying factors over which principals exert minimal control, leading to biased estimates of individual principals’ effectiveness and an overstatement of the magnitude of principal effects. Based on our framework, which better accounts for bias from time-varying factors, we find that little of the variation in student test scores or attendance is explained by persistent effectiveness differences between principals. Across contexts, the estimated standard deviation of principal value-added is roughly 0.03 student-level standard deviations in math achievement and 0.01 standard deviations in reading.
Prior research has found that economic downturns have positive effects on new teacher quality, but has not been able to determine the extent to which this relationship arises from a supply response (increased quantity or positive selection of teaching candidates) vs. a demand response (selection in hiring enabled by falling demand). In this paper, I use longitudinal data on students and teachers in Massachusetts to describe the effects of higher unemployment rates on both supply and demand for teachers. I show that students who graduate from college when unemployment rates are higher are more likely to take a teacher certification test, and that this effect is stronger among students who were higher achieving while in high school. On the demand side of the market, higher unemployment reduces new teacher hiring and the overall number of teachers employed, but I find no evidence that schools differentially employ higher achieving teaching candidates during economic downturns. While I cannot definitively rule out changes in demand-side selection, I show that much of the positive relationship between unemployment rates and teacher quality can be explained by positively selected supply. My results suggest that economic incentives impact both the quantity and the quality of new teaching candidates, with implications for attracting and retaining high-quality teachers outside of economic downturns.