33760 research outputs found
Sort by
Spiritual Equality, Social Hierarchy: Gender and the Construction of Quaker Identity in The Early Anglo-American World
No full textThis thesis is a study of how early Quaker doctrines and religious ideas influenced Quaker gender identity in the Anglo-American colonies in the late seventeenth and early eighteenth centuries, mostly between 1660 and 1720. The Quaker concept of the Inner Light undermined the spiritual hierarchy that informed the worldly social structures of early modern England and New England. This was true both within and outside of Quaker communities. Quaker women found vastly expanded social roles through the Inner Light, including preaching, publishing, and itinerant ministry but maintained a strong sense of physical and mental gender difference. Quaker men redefined masculinity to better reflect their view of Christ-like manhood while still retaining social patriarchal structures. Quakers largely welcomed the social changes that came with spiritual equality, but New England Puritans were threatened by its radicalism and worried that the Inner Light undermined their power over the new colony. The flexibility of Quakerism in its earliest years allowed for a redefinition of gender identities, but seventeenth century Anglo-American patriarchy was not so flexible that Quakers rejected or considered rejecting gender norms completely
A FRAMEWORK FOR DATA COLLECTION AND AIRSIDE OPERATION METRICS ANALYSIS AT SMALL AIRPORTS
No full textWith the growing global demand for air travel, General Aviation (GA) airports are facing significant challenges. Unlike larger airports, many GA airports operate with limited infrastructure, leading to issues such as delays and congestion. Managing a mix of flight types, including training and regular flights, within tight budget constraints and limited runway capacity further complicates operations. Effective management and reliable capacity estimation are crucial, especially as these airports often depend on federal funding for future expansions. However, the lack of effective data collection mechanisms and equipment makes it difficult to implement data-driven management strategies or accurately estimate capacity, particularly given the complexities of handling diverse flight operations.
Tasked by the Federal Aviation Administration (FAA), this project addresses the capacity estimation challenges at GA airports using Automatic Dependent Surveillance–Broadcast (ADS-B) technology. It proposes a comprehensive data pipeline and analysis system hosted on Amazon Web Services (AWS) to collect, decode, filter, analyze, and archive flight data. This system facilitates the extraction of key operational metrics for advanced capacity modeling. To ensure precise parameter extraction, the framework incorporates a rule-based model for accurate operation type classification. Additionally, a novel signal enhancement method is introduced to improve ADS-B data quality, ensuring more reliable and consistent flight trajectory timestamps.
To support the development of the second generation of the Airport Capacity Model (ACM2) and define the required operational metrics, this work provides specifications for bounding boxes at target airports and establishes key operational benchmarks. The methodologies for calculating departure and arrival operational metrics based on the benchmarks are also detailed.
Leveraging the advantages of the proposed data analysis system, this study demonstrates various applications of ADS-B data analysis. These include performance comparisons between flights with different operational purposes, correlations between squared flight speeds at various phases and density altitude, and time series predictions of air traffic flow at specific airports. By addressing these challenges, this project has the potential to significantly enhance the accuracy of capacity estimation across thousands of GA airports while delivering reliable aviation data and actionable insights to both the aviation research community and GA airport stakeholders
GRAPPLING WITH THE IMPLEMENTATION OF READING PRACTICES: EDUCATORS CROSSING BOUNDARIES IN A MULTI-SCHOOL / UNIVERSITY RESEARCH-PRACTICE PARTNERSHIP
No full textResearch Practice-Partnerships (RPP) offer an organizational structure and methodological framework for researchers and practitioners to learn together to use research to inform school and classroom systems and implement evidence-based instructional practices to increase positive outcomes for students (Diamond, 2021; Estrada & Tanksley, 2022). As the number of RPPs increases and the body of research grows more expansive and nuanced, empirical studies with finer-grained analysis are important for the field to understand how RPPs support learning between researchers and practitioners to address educational challenges. This study investigates an RPP between a public school district and a school of education at a private university through an in-depth examination of how boundary-spanners, members of the RPP organizations who are responsible for connecting the organizations’ practices, learn together to support the implementation of evidence-based instructional reading practices in K-12 elementary schools as well as research and learning within the university’s teacher education program.
Qualitative single-case study methodology and analysis were used for this instrumental case to examine how boundary-spanners learned together to further the goals of the RPP. Boundary-spanning and RPP organizational frameworks guided qualitative analysis, which included an exploration into how boundary-spanners addressed equity in the partnership (Akkerman & Bruining, 2016; Farrell et al., 2022; Sjölund, 2024; Yamashiro et al., 2023). Data collection included observation of RPP leadership meetings, monthly professional development sessions with school leaders, interviews with RPP leaders, school leaders, and university personnel, and documents from the RPP researcher and District teams.
Findings revealed that boundary-spanners learned together at specific boundaries related to the differing expectations of when and how school leaders should implement practices, the capacity of school leaders to plan for implementation, and the alignment of practices to the District curricula. This analysis also revealed that specific issues of outcome equity were not a place for boundary-crossing with this team. The RPP university team also engaged with teacher education faculty and staff to share learning from the RPP despite not setting specific goals for university transformation. An unexpected finding emerged when the partnership’s future was jeopardized because funding would no longer be available. This prompted District RPP leaders to make strategic decisions to ensure the RPP’s continuation.
Based on these findings, recommendations are made for expanding current RPP conceptualizations to include an indicator of each organization’s readiness for addressing equity issues within the absorptive capacity and, within boundary infrastructure, an explicit goal for research entities’ transformation. This fine-grained analysis of boundary spanners’ learning and examination of one organization’s fluctuating capacity to learn contributes to the RPP field’s understanding of what happens within an RPP’s boundary infrastructure to support changes in practices that could lead to educational transformation for both schools and universities
Art to Support Community Goals: Takoma Langley Crossroads
No full textFinal report for ARTT426: Advanced Painting: Painting on Site (Spring 2025). University of Maryland, College ParkThe students in this course, led by Professor Brandon Donahue-Shipp at the University of Maryland, collaborated with the Takoma Langley Crossroads Development Authority (TLCDA) to create a public art and beautification initiative focused on sustainability and community engagement. The project involved repurposing old street planters along New Hampshire Avenue and University Boulevard through custom mosaic tile artwork. Guided by site visits, community engagement, and a hands-on workshop with local mosaic artist Graciela Granek, students developed designs rooted in the cultural and aesthetic identity of the neighborhood. The culmination of the project featured a public installation and event in early May, where students displayed their work and facilitated interactive art activities. The project served as a platform for creative placemaking, blending environmental consciousness, community voice, and artistic practice in a highly visible public space.Montgomery County, MDhttps://vimeo.com/1090605839/f6e178300c?share=cop
ACCURATE AND SCALABLE PHYLOGENY ESTIMATION VIA GRAPH CUTS
No full textReconstructing evolutionary relationships among populations or species (called a species tree) is an important precursor of many biological studies, with applications in agriculture, medicine, and conservation. A major obstacle to species tree reconstruction is that the evolutionary histories of species can vary across the genome due to incomplete lineage sorting, a biological process modeled by the Multi-Species Coalescent. Many of the leading methods developed to date, for example, the ASTRAL family of methods, reconstruct the species tree from gene trees (i.e., a tree built from the related regions of the genomes across different species). ASTRAL address heterogeneity by evaluating evolutionary relationships on four species at a time, as these quartets have favorable statistical properties under the Multi-Species Coalescent. Despite significant advances, species tree reconstruction continues to be challenged by the complexity of evolutionary processes, the hardness of related optimization problems, and the quality of data; thus, new algorithms are needed. This dissertation presents advanced methods based on graph cuts, yielding improvements in scalability, accuracy, and robustness.
First, I introduce TREE-QMC, a heuristic for the NP-hard Maximum Quartet Support Species Tree problem. TREE-QMC is based on the divide-and-conquer approach proposed by Snir and Rao (2010). My main contribution is showing how to build an object called the quartet graph directly from gene trees, without explicitly enumerating all quartets. This result enables TREE-QMC to be cubic time in the number of species, making it the first method to break the quartic time barrier without down-sampling the input quartets since the divide-and-conquer framework was proposed. Second, I introduce the notion of a normalized quartet graph and show how to efficiently integrate normalization into graph construction. Together, these contributions enable TREE-QMC to achieve greater accuracy and scalability than ASTRAL-III, the leading method at the time, on data sets with large numbers of taxa (500-1000) and other challenging conditions.
Second, I reformulate the TREE-QMC algorithm to weight quartets based on gene tree branch lengths and support values, as proposed by Zhang and Mirarab (2022). Although the weighting scheme improves robustness of TREE-QMC to poor quality inputs (i.e., gene trees with missing species and/or estimation error), it comes with a small increase in time complexity compared to the unweighted algorithm. Fortunately, the increase in running time is small in practice, behaving more like a constant factor. Moreover, weighted TREE-QMC is highly competitive with weighted ASTRAL-IV, the leading method at the time, again producing more accurate species trees on data sets with large numbers of taxa (500-1000) and other challenging conditions.
Third, I demonstrate the utility of TREE-QMC for evolutionary scenarios, like hybridization, where the species history is a network rather than a tree. Recent research shows that reconstructing the tree-like aspects of a network, called the tree of blobs, is important for scalable network reconstruction via divide-and-conquer. An obstacle here is that the leading method for tree of blob reconstruction, TINNiK, does not scale to large numbers of species. To address this issue, I propose to build a tree with TREE-QMC and then contract edges in it based on statistical testing. This approach enables greater scalability and accuracy than TINNiK, especially on data sets with high amounts of incomplete lineage sorting.
Overall, the algorithms presented in this dissertation advance species trees and tree of blob reconstruction and highlight avenues for future research
QUANTIFICATION AND ANALYSIS OF SPATIO-TEMPORAL WAVES IN DYNAMIC CELLULAR SYSTEMS
No full textRhythms are signatures of life. From the nanoscopic flicker of molecular switches to the metronomic beating of the heart, internally generated clocks regulate virtually every aspect of physiology, thereby shaping the very essence of living systems. This dissertation addresses two interrelated questions: How can such biological rhythms be captured and quantified with sufficient resolution, and do common design principles underlie their organization across different biological scales? This work follows a cohesive narrative that investigates rhythmic phenomena across progressively smaller spatial and temporal domains, while concurrently developing novel analytical tools to advance their characterization.
We begin with a survey of oscillatory phenomena spanning multiple orders of magnitude, introducing multiscale modeling as a unifying framework for integrating diverse biological rhythms. With this broad context established, we turn to specific experimental systems, starting with gut motility in an ex vivo crayfish model. By isolating central, myogenic, and serotonergic inputs, we find that chemical cues can restore contraction strength but reduce spatial synchrony in the absence of central control. These results suggest that local sensing mechanisms may be essential for coordinating large-scale motor patterns.
The focus then shifts to a finer spatial scale: actin dynamics in astrocytes, a type of glial cell. Using a custom optical-flow analysis pipeline, we identify recurrent actin "hotspots" whose activity is suppressed by engineered nanotopographies but enhanced in the presence of neighboring neurons. These findings suggest that the cytoskeleton itself functions as a dynamic sensor of mechanical and biochemical cues. Building on this insight into cellular sensing, we next examine how cells interact with their environment during development. In growing cortical neurons, actin wave tracks and growth-cone trajectories initially align with nanotopographic cues; however, this influence diminishes as axons mature. The observed decline points to age-dependent cytoskeletal plasticity as a potential factor limiting regenerative capacity in adult neurons.
In the final part of this work, we extend our analysis from two-dimensional imaging to the three-dimensional microenvironments in which cells naturally reside. As most, if not all tissues function in three dimensions, spatial depth fundamentally alters how motion is encoded and perceived. To address this complexity, we adapt the optical-flow framework to volumetric datasets. Applications to actin dynamics in Dictyostelium and axonal growth in Drosophila pupal wings reveal spatial patterns obscured in two-dimensional projections. Complementary experiments involving calcium imaging and electrophysiological recordings in the gut further link local intracellular activity with large-scale contractions, thereby reinforcing the broader theme of wave propagation across scales.
Overall, this work aims to bridge wave phenomena across diverse spatial and temporal domains, unifying them under the central theme of local versus global control in the coordination of biological rhythms
Bridging the Gap Between Biomechanics and Modern Computer Architecture: Analyzing Bottlenecks in a Human Eye Simulation
No full textFinite element analysis (FEA) is a crucial tool in biomechanical simulations, enabling the study of complex biological systems under various conditions. FEBio, an open-source FEA software designed for biomechanic and biomedical research, is widely used for such applications. However, its computational performance across various workloads remains an important consideration for researchers seeking efficiency and scalability. This thesis evaluates the performance of FEBio using system-level and microarchitecture metrics, profiling tools, and comparative analysis in biomechanical simulation workloads. A case study on a finite element model of the human eye, from the glaucoma treatment study that inspired this project, provides a practical demonstration of FEBio’s computational behavior under realistic conditions. By identifying performance bottlenecks, analyzing their underlying causes, and proposing potential optimizations, this research contributes to improving the efficiency and resource utilization of biomechanical simulations, furthermore facilitating fruitful studies in the fields of bioengineering
ESSAYS ON MIGRATION AND DEVELOPMENT ECONOMICS
No full textThis dissertation studies how large-scale population movements, whether through migration or colonization, shape economic outcomes in receiving economies. It focuses on their effects on native workers, firms, local markets, and trade-competitive advantage.
The first two chapters examine contemporary immigration shocks and highlight different channels through which immigration influences host economies. Chapter 1 investigates the effect of immigration on firms’ capital deepening, focusing on the role of immigrant skills. We exploit the quasi-natural experiment of the arrival of one million immigrant workers from Venezuela and Haiti to Chile between 2015 and 2019, which generated substantial variation in labor supply by skill level. Using a unique administrative dataset that links workers to firms across the entire Chilean private sector, combined with confidential balance sheet data, we find that a one percentage point increase in the share of skilled immigrants among a firm’s skilled workforce leads approximately to a 2 percent increase in capital per worker. This relationship is nonlinear: the effect diminishes as the firm’s share of immigrants rises, consistent with imperfect substitutability between skilled natives and skilled immigrants. The results highlight the potential for immigration to stimulate firm-level investment and suggest that the effects of immigration are stronger when labor market integration between immigrants and natives is high.
Chapter 2 evaluates the impact of immigration shocks on local housing markets in Italy, an important but less studied channel through which immigration can affect native welfare. Housing prices and rents influence the real income of native residents and can trigger native relocation, thereby affecting broader labor market dynamics. I estimate the Italian housing market’s response to increased immigration from 2010 to 2019, exploiting geographical and temporal variation in immigrant settlement patterns. Theoretical predictions are ambiguous: while demand-side pressures should raise prices, native outflows could exert downward pressure. The empirical results show a positive and statistically significant association between immigrant inflows and both housing prices and rents. On average, a 1 percent increase in the immigrant stock raises housing prices by 0.03 percent and rents by 0.04 percent. I find no evidence of native flight.
Chapter 3 turns to a historical context and develops a theoretical model to explain the shift in comparative advantage in the textile sector between India and Great Britain during the colonial period. We analyze the role of an exclusive trading company that monopolizes the foreign trade of a colony, using the East India Company (EIC) as a case study. The framework explains both the rise and the eventual dismantling of the EIC, particularly in light of Britain’s Industrial Revolution. More broadly, the chapter suggests that colonial policies might be particularly burdensome for colonies with rival economies, a situation often overlooked in studies of the New World, where colonies were typically complementary to their metropolises
ON LEARNING BEHAVIORS OF PARALLEL CODE AND SYSTEMS ACROSS MODALITIES
No full textPerformance modeling is an integral part of the research process forcomputational scientists. It enables them to understand how different factors
contribute to the final runtime of an application. This understanding is crucial
to developing efficient scientific applications and simulations. While
important, performance modeling is difficult as there are a large number of
factors that may contribute to final performance. Factors such as the algorithm,
problem size, implementation, architecture, and systems software stack all
impact performance in an often complex relationship. Analytical models can be
employed to study these causal variables and performance, however, they are
difficult to scale up to a large number of input variables. Additionally, the
relationship between the causal variables and performance may be unknown or
complex, making it challenging to derive an analytical model. Fortunately,
machine learning (ML) can help address these challenges as ML algorithms excel
at modeling unknown and complex relationships. Furthermore, ML-based performance
models can handle a large number of input variables, making them ideal for
modeling complex scientific codes. By training ML models on historical
performance data, computational scientists can develop accurate models that can
predict the performance of new applications and simulations under different
scenarios. However, current ML-based modeling approaches are limited to modeling
one or two sources of performance data, such as hardware counters or application
features. This limitation prevents models from making use of all available
causal variables that may impact performance. This thesis introduces novel
approaches to modeling performance that can make use of all available data
sources. Additionally, it introduces performance latent spaces that can be used
to model various output metrics, such as runtime or energy consumption, in a
unified manner. Finally, a method to integrate these performance models into
large language models is introduced to enable modeling and improving the
performance of code
ANALYTIC ACTIONS ON CLOSED, CONNECTED -DIMENSIONAL MANIFOLDS
No full textThis thesis provides a classification of analytic actions of the semiorthogonal group , where , on closed, connected -dimensional manifolds. Adapting Uchida's construction of actions on , we explicitly construct analytic actions of on and , as well as actions on , where is a maximal parabolic subgroup of . The central result of this thesis demonstrates that any analytic action on a closed, connected -dimensional manifold is covered by one of the constructed actions. For , the actions of correspond to a particular class of vector fields on the circle, while for , they correspond to actions of on either the sphere or the torus