71446 research outputs found
Sort by
The Digital Decline of Social Capital: How Social Media Amplifies Political Polarization And Why Better Tech is Not the Answer.
No full textPolitical polarization in the US is at an all-time high with some social scientists hypothesizing that the United States has reached a level not seen since the Civil War. Prior research points to multiple contributing factors including the decline of social capital. Social capital, as argued by Robert Putnam in his seminal work, Bowling Alone, is required to maintain individual social bonds, support broader community engagement and reinforce strong democracies. However, with the rise of digital technology, and more specifically, the growth of online social networks in the last 25 years, social capital has been declining, not increasing.
This paper will show how these social networks (SNs) have been unable to reproduce the types of networks, norms and trusts that Putnam outlines are required in order to build and maintain social capital while also exploring how these SNs and the algorithms that power them, work to spread mis- and disinformation, surface conspiracy theories and normalize hate speech, all of which have contributed to a rapid rise in political polarization. The paper will also show how these platforms have become a primary source of news and information for the majority of Americans, thereby eliminating trust between users and groups and ultimately reshaping the legacy media landscape. Finally, this paper will examine the critical role these platforms have played in various political campaigns since 2000 and how both major parties have leveraged SNs to fortify epistemic bubbles and echo chambers in an effort to build deep trust within their
base (in-groups) while also fostering extreme feelings of anger, distrust and apathy towards supporters of the opposing party (out-groups).
The long-term impact of SNs on broader attitudes towards both in-groups and out- groups following the 2024 election continues to evolve but in looking at prior scholarship, there is a need to find alternatives to technological solutions if society hopes to tamp down the fires of extreme political polarization and rebuild the nation’s dwindling supply of social capital.Extension Studie
Efficacy of an Online, Group-Based Internal Family Systems (IFS) Intervention on PTSD Symptom Clusters
No full textBackground and Objectives -
Posttraumatic stress disorder (PTSD) is a debilitating condition that often follows
exposure to traumatic experiences. Its symptoms are defined in the Diagnostic &
Statistics Manual, 5th edition, text revision (DSM-5-TR; American Psychiatric
Association, 2022) and organized by specific criteria (intrusion Criterion B, avoidance
Criterion C, negative cognition Criterion D, hyperarousal Criterion E). Current treatments
show promise in alleviating PTSD symptoms but rarely fit all patients. There is,
therefore, a need for continued development of new modalities. Internal Family Systems
(IFS) is an innovative intervention that has gained popularity since its inception forty
years ago. Preliminary studies have shown its potential in alleviating symptoms of
traumatic stress, yet, evidence of its efficacy remains to be established. This individual
participant data (IPD) meta-analysis aims to scrutinize the efficacy of an online, group
based IFS intervention on PTSD symptom clusters. It postulates that overall PTSD
symptoms will improve, with greatest reductions seen in intrusion (Criterion B) and
avoidance (Criterion C) symptoms.
Methods -
Four studies (N = 76 participants) were included in this IPD meta-analysis. All
involved an online, group-based IFS intervention for PTSD paired with individual IFS
therapy sessions. Pre-/post-intervention PTSD CheckList for DSM-5 (PCL-5; Blevins et
al., 2015) data were collected via self-reported surveys. The primary aim of this work is
to assess the efficacy of the IFS intervention by cluster of PTSD symptoms as defined by
the DSM-5-TR. A two-stage individual participant data (IPD) meta-analysis with random
effects, measuring differences in effect sizes followed by meta-regressions with and
without moderating variables were performed.
Results -
The meta-analysis by cluster showed large negative effect sizes for all symptom
clusters (gB= -1.17, gD= -1.12, gC= -1.04, gE= -0.80, p.05, pHB 0.05), yet with low
certainty around the estimates (wide 95% CIs). Heterogeneity statistics did not reveal any
between-study variability, indicative of consistent study protocols and intervention
designs. While intrusion (Criterion B) and negative cognition (Criterion D) appeared to
have markedly higher effect sizes, a meta-regression using clusters as a moderator,
disproved the significance of the effect size differences (p>0.05, pHB > 0.05). Hence,
clusters do not contribute differently to the reduction in PTSD symptoms.
Discussion and Conclusions -
Overall, the IFS intervention showed significant efficacy in reducing overall
PTSD severity. Even though efficacy was not linear across PTSD clusters, and intrusion (Criterion B) and negative
cognition (Criterion D) in particular showed observable reduction in gravity, no cluster significantly differed
in terms of effect sizes. These results indicate that IFS improves symptoms of PTSD
homogeneously across symptom clusters. This aligns with empirical findings from the
existing literature regarding other PTSD treatment modalities (including CBT and
EMDR), which show a lack of differentiation in efficacy amongst PTSD symptom
clusters. All in all, this IPD meta-analysis adds to the emerging research showing the
efficacy of IFS as a modality of psychotherapy to alleviate PTSD. It is encouraging as
novel types of therapy are needed to complement existing interventions and support
patients who do not benefit from them.
Scientific Significance -
An innovative modality of psychotherapy, IFS could effectively alleviate
symptoms of PTSD. The results of this IPD meta-analysis carry important implications
for people suffering from PTSD. Adding a novel, evidence-based option for therapy that
has the potential to holistically improve PTSD symptoms is encouraging for the many
people who do not find relief with other existing modalities of treatment.Extension Studie
Essays on the Economics of Information: Gender and Innovation
No full textThis dissertation examines how agents process information and how it diffuses through the economy, with particular attention to the roles of gender and technological innovation. Chapters 1 and 3 study how gender shapes the interpretation of information about firm performance. Chapter 2 focuses on the diffusion of novel technologies across geography and labor markets. Together, the three chapters contribute to our understanding of the economics of information by showing how identity and context shape the way agents interpret and act on new information.Economic
Sold to the Highest Bidder: Can We Rethink Fairness in Targeted Advertising, or Are We Placing a Losing Bet?
Get PDFThe pursuit of fairness in algorithms is often framed as a solvable mathematical problem, yet the concept of fairness itself has long been debated by philosophers, historians, and economists without resolution. If defining fairness were merely a statistical challenge, then computer scientists might have succeeded where others have failed. However, the real difficulty lies not in measuring fairness but in determining whose perspective defines it. Fairness is not an inherent property of an algorithm; rather it is a function of the relationship between the algorithm and the stakeholders involved.
This thesis examines fairness in online advertising auctions, where algorithmic decision-making determines which users see which ads. Instead of treating fairness as a fixed standard, this work models different fairness definitions and evaluates their impact on ad distribution, efficiency, and stakeholder incentives. The modeling framework presented here does not seek to prescribe an "optimal" fairness solution but instead reveals how different fairness constraints shape online advertising outcomes, exposing trade-offs between fairness, efficiency, and revenue. By simulating ad auction environments under competing fairness criteria—such as statistical parity, equalized odds, and individual fairness—this thesis demonstrates how small changes in fairness definitions can yield vastly different systemic effects.
Rather than positioning fairness as a purely computational objective, this work argues that fairness in ad auctions is fundamentally a policy decision, not a technical optimization problem. Through this modeling, it becomes evident that decision-makers—whether policymakers, companies, or regulators—hold the power to define and enforce fairness, shaping market outcomes through their chosen definitions. This thesis challenges the assumption that fairness can be universally quantified, instead advocating for transparency in the trade-offs that fairness interventions impose on different stakeholders. Algorithmic fairness is not about finding a single solution, it is about making explicit the hidden values embedded in digital decision-making and understanding that any notion of algorithmic fairness is ultimately a reflection of the human mind behind the machine: defining, building, and ultimately, making a decision.Applied Mathematic
Essays in Bayesian Econometrics
No full textThis thesis comprises three chapters in Bayesian econometrics. Although the chapters are independent,
the unifying theme is that Bayesian inference is semiparametric. That is, the econometric
model is either 1. indexed by finite- and infinite-dimensional parameters, or 2. interest is a finite-dimensional
transformation of an infinite-dimensional parameter. All chapters are solo-authored.
Chapter 1 proposes Bayesian inference for conditional moment equality models. The framework
starts with a prior for a conditional distribution and reports a marginal posterior for an estimand
that minimizes the distance of the conditional moments to zero. The key theoretical result is a
Bernstein-von Mises theorem, establishing asymptotic normality of the minimum distance posterior.
Chapter 2 presents a new approach to Bernstein-von Mises theory for partially linear regression
models. The idea is to embed an adaptive parametrization of the regression function within a
(quasi-)likelihood model, enabling verification of the Bernstein-von Mises theorem using ordinary
expansions of the log-likelihood function. The new parametrization alleviates some smoothness
restrictions that are encountered in the original parametrization of the model.
Chapter 3 introduces a Bayesian inference framework for a linear index threshold-crossing binary
choice model subject to a median independence restriction. The proposal exploits an observational
equivalence between the model and a probit model with nonparametric heteroskedasticity. This
leads to a computationally attractive Bayesian inference procedure in which a Gaussian process
forms a conditionally conjugate prior for the natural logarithm of the skedastic function.Economic
Electroadhesion Design for Microrobotic Locomotion on Diverse Surfaces
No full textSmall bioinspired insect-scale microrobots have the potential to someday be useful for exploration or inspection tasks, with their small size allowing for access to confined spaces. However, real-world inspection tasks often involve surfaces of variable inclines and finishes, highlighting the need for small-scale robust climbing options that are adaptable to diverse terrains. In this work we examine development of the Harvard Ambulatory Microrobot (HAMR), a 4.5 cm, 1.5 g microrobot. Improvements to the fabrication process for more robust manufacturing are made, with demonstration of an even smaller scaled (22.5 mm, 320 mg) HAMR-JR robot.
Electroadhesion improvements and robot capabilities are further explored for the purposes of enhancing locomotion and payload capacity over a diversity of pristine, rough, and inclined terrains. On pristine, smooth, flat terrains, improvements using thinner dielectric layers enable fabrication of electroadhesive pads capable of 60 g of shear force (roughly 40 times the weight of a robot). On these smooth terrains, large simple circular foot pads exhibit the greatest shear forces. However, on rougher inclined surfaces, pads which adjusted the width, length, and number of spoke-like features provide greater compliance and achieve more consistent shear adhesion forces. The improved adhesion capabilities of a compliant spoked foot pad are demonstrated with enhanced robot locomotion over steeper 37 degree inclines on a rough (75 µm vertical and 1 mm horizontal spatial roughness) conductive surface. Further design explorations of various compliant kirigami geometries are conducted to consider spoke, serpentine, and fractal patterns, with a simple model developed for spoke design parameters. Serpentine designs are found to most enhance compliance and shear adhesion capability on rough surfaces (up to twice the adhesion force of circular designs). Several demonstrations of the robot using compliant electroadhesive pads illustrate the potential of the robot to be used in industrial environments, with simple manipulation and locomotion tasks over a variety of conductive rough, inclined, and curved terrains.Engineering and Applied Sciences - Engineering Science
Reclaiming History: Black Lives at Sweet Briar College
No full textSweet Briar College was founded in 1901 on the land of a former Virginia plantation and was constructed and run by many of the descendants of the enslaved. As recently as 2019, it was estimated that thirty percent of the hourly wage workers of the college were descendants of the enslaved. While many universities and colleges have been reckoning with their slave past, Sweet Briar has lagged far behind in this self-examination. My thesis, through interviews and research, attempts to understand how the history of the family that owned the plantation influenced the way Black lives at the College were marginalized and why only one of two U.S. colleges built on the land of a former working plantation has done so little to examine its history, even at a time when many other colleges and universities are engaged in that research. The College promoted itself as a place where the ideals of Southern white womanhood were elevated, preserved and perpetuated and the racial divide between white students and Black staff was sharply defined. This thesis also looks at those who, though often in the minority, attempted to push back against the administration and Board during critical times of change, such as the period of integration and currently, as Sweet Briar has changed its admissions policy to bar transgender applicants.
I continue the work of historian Lynn Rainville, a professor who first began to research the college’s slave history in 2001. Rainville is the only academic who has studied the founding of the college in a full light, illuminating the false narrative rife with the Lost Cause paternalism that had previously been the published and accepted history of the school.
In my examination of the myths of the Lost Cause in the South and its influence on generations of white students at the college and their families, this thesis attempts to see the influence false ideas about race, privilege and social order have had on the operation of the college. I attempt to gain an honest understanding of how the College handled matters of race, interviewing former Black and white students from the period of the 1960s and 1970s to see how this racially imbalanced environment affected their experience as students. The college archives and school publications from the earliest periods and throughout the mid-to-late 20th century also feature significantly in my research in their ability to reveal student life at the College as well as the actions of the administration and the Board regarding many aspects of life at Sweet Briar. It is my aim in this research to further the work that has begun in this area that could move Sweet Briar College forward in a racial reckoning and a more honest and fulsome history.
A key finding of my research reveals that the will of the former slave owner that created the College has been interpreted in conflicting ways by the Board at pivotal times in its history to justify controversial changes that they wished to make. This practice is continuing currently with the College’s decision in 2024 to change the admissions policy to bar transgender women based upon a 1899 definition of “woman.” It was an unexpected change in policy for a college that currently has transgender students and thirty percent of the student body identify as nonbinary. In the current political climate in the U.S., it is difficult to project what effect this will have on the future of Sweet Briar College. The history of the school I have researched and written about suggests that facing the current realities of life is not something that comes easily or quickly.Extension Studie
Deep Generative Models for Prediction and Design of Enzymes
No full textOver billions of years, proteins have evolved functions that drive nearly all biological processes on Earth. This vast evolutionary record offers an enormous experimental dataset that enables predictive modeling of biological systems. In this thesis, I establish benchmarks for protein variant effect prediction and novel sequence generation using machine learning (ML). I present a series of applications that reveal strengths and limitations of ML in minimizing experimental efforts in protein design and human genetics, opening new avenues for biological discovery.
First, we develop benchmarks for both prediction and generation of enzymes. In Chapter \ref{chap1}, we establish the current state of the art by benchmarking over 40 machine learning models against 250+ experimental datasets, offering the most comprehensive evaluation of protein design models to date. Chapter \ref{chap2} proposes benchmarks for sequence generation; we investigate TEV protease as a case study to evaluate the generative capacity of ML models in designing novel protein sequences. Testing over 100,000 variants for both expression and protease activity, we illuminate the biological consequences of different modeling approaches and provide insights into generative design strategies.
Then, we focus on applications to three proteins: a gene editing enzyme called RfxCas13d, a subunit of an amino acid synthase called Tryptophan Synthase Beta Chain (TrpB), and a neuron-specific protease called Botulinum Neurotoxin (BoNT). Chapter \ref{chap3} focuses on RfxCas13d, a CRISPR enzyme capable of both \textit{cis} and \textit{trans} RNA cleavage, for which limited natural sequences and no structural data exist. By developing a novel ML-guided approach, we nominate experimental positions and achieve a 7-fold improvement in the enzyme's targeted property. In Chapter \ref{chap4}, we extend our analysis to enzymes like TrpB and BoNT where sequence-based data alone proves insufficient for designing new functions. We underscore the limitations of current unsupervised learning approaches and emphasize the necessity for alternative, data-integrative modeling techniques. Finally, Chapter \ref{chap5} explores the development of new generative models to predict human genetics, both in coding regions and in non-coding regions of the human genome.
By establishing clear benchmarks and designing novel proteins, this work not only advances the field of protein design but also lays the foundation for the next generation of models in both variant effect prediction and novel sequence generation.Medical Science
Deorphanizing Enzymes: Characterization of NXPE1 and Its Role in Mucin Glycosylation and Ulcerative Colitis Pathogenesis
No full textOrphan enzymes represent a largely unexplored frontier in molecular biology, with their unknown substrates and functions posing significant challenges and opportunities for understanding human health and disease. This thesis investigates the biochemical role of NXPE1, a gene implicated in ulcerative colitis (UC), to uncover its enzymatic function and physiological relevance. NXPE1 encodes an O-acetyltransferase that selectively modifies the sialic acid N-acetylneuraminic acid (Neu5Ac) at the 9-OH position, producing Neu5,9Ac2. This acetylation process, localized to the Golgi apparatus of colonic epithelial cells, is critical for modulating the biophysical properties of mucus glycoproteins. Using a multidisciplinary approach that integrates in silico modeling, activity-based protein profiling (ABPP), and organoid-based in vitro validation, this study demonstrates that the UC-protective variant NXPE1 G353R disrupts enzyme stability and acetylation activity.
Key findings reveal that NXPE1-mediated sialic acid acetylation is essential for maintaining the structural and functional integrity of the colonic mucosal barrier. The absence of NXPE1 or loss of function due to genetic variation correlates with increased mucus viscosity and altered mucosal sialoglycome, which may protect individuals from inflammation and dysbiosis. These findings not only elucidate a mechanistic basis for the protective effect of NXPE1 variants in UC but also highlight the broader significance of sialic acid modifications in health and disease. By deorphanizing NXPE1, this thesis advances our understanding of enzyme function and its implications for therapeutic strategies targeting inflammatory bowel diseases.Chemical Biolog
The Discrete Bootstrap for Quantum Thermodynamics
No full textA novel bootstrap method is exemplified by lattice theories and quantum mechanics. Positivity of expectation values combined with sets of non-trivial constraint equations and semidefinite optimization rigorously and precisely bound correlation functions. Starting with the example of a one-dimensional integral, the technique is built upon to extract limit cycles in a two-dimensional classical dynamical system and spin correlation functions in the Statistical Ising Model. Further positivity constraints and simplifications arising in a large rank limit are implemented to bound the energy expectation value as a function of temperature in Matrix Quantum Mechanics. Along the way, the S-matrix of Chern-Simons-matter theory is computed as an example of computational simplification at large gauge group rank.Physic