18693 research outputs found
Sort by
Do Designated Market Makers Facilitate Earnings News Discovery?
No full textAs markets replace contractual liquidity providers (designated market makers; DMMs) with voluntary liquidity provision through cutting-edge technology, we investigate how this affects price discovery. Research suggests that endogenous liquidity provision is not always optimal. We investigate how DMMs affect the incorporation of earnings news into prices. Using a regression discontinuity design, we show that increased DMM participation facilitates earnings news discovery—lower JUMP, lower Synchronicity, and higher Future Earnings Response Coefficient. Greater DMM participation associates with improved liquidity, and induces greater informed trading as evidenced by more short selling on negative news and increased EDGAR and Bloomberg search activity before earnings announcements. Our results highlight an important and hitherto overlooked effect of modern technology on processing earnings information
A Framework for Vehicle Bridge Strike Detection, Strike Characterization, and Damage Estimation for Railroad Bridges
No full textRail bridges provide vital crossings for freight and passenger trains over natural and man made obstructions in terrain. Over time these structures develop damage due to aging or unexpected external loading events. Specifically, low clearance rail bridges are susceptible to frequent strikes from overheight vehicles or equipment. It is critical to detect those strikes once they occur to ensure the bridge and the public safety and to also meet FRA regulations of mandatory post-strike bridge inspection. Early bridge damage detection also reduces bridge closure times and prevents further deterioration. Not every bridge strike represents an immediate risk to safe bridge operation, thus this dissertation presents a comprehensive methodology that detects vehicle bridge strikes in real-time, characterizes strike severity post detection, and detects and quantifies damage if present in the bridge. First, the system leverages the improved accessibility and scalability of bridge instrumentation technology and interrogates bridge data using mechanics and Machine Learning (ML) algorithms to rapidly detect strikes and determine whether an immediate inspection is necessary or can be safely deferred. Specifically, this dissertation develops parallel heterogeneous data-fusion convolutional neural networks (PHD-CNN) operating on data collected from in service rail bridges to improve detection and classification of vehicle-bridge strikes. The method provides a mechanism to homogenize and fuse disparate data streams for use as inputs to a classifier that distinguishes bridge strikes from passing trains. Optimum PHD-CNN networks detect, on average, 95% of bridge strikes with false positive rates less than 2%. Next, operating on identified strike data the framework utilizes principal components analysis, an unsupervised machine learning technique, to characterize strike severity. The system analyzes extracted v severity-related features to group strikes with similar characteristics together and then compares them to user defined thresholds to determine strike severity. Finally, for an observed change in the system’s fundamental frequency this dissertation presents an energy-based mechanics relationship to provide a feasible domain of potential damage scenarios to detect, localize, and characterize damage. The final output of the system comprises practical guidance to inspectors by (1) indicating the presence of damage, (2) locating the damage, and (3) quantitatively estimating the severity of the damage; thus, the method attains a Rytter level 3. Rytter levels comprise four stages of damage evaluation: detection, localization, quantification, and prediction of remaining structure life. They are a widely used framework in structural health monitoring to rate the capabilities of damage assessment systems
Defense Use of Digital Discovery in Criminal Cases: A Quantitative Analysis
No full textRecent criminal court reforms have required prosecutors to provide defense attorneys with broader and earlier discovery of evidence. For these discovery reforms to fulfill their aims of improved fairness and efficiency, defense attorneys must take advantage of the evidence disclosed by the prosecution. Prior studies suggest, however, that a range of factors, including low pay and high caseloads, impede effective defense representation in general. If similar factors hinder defense attorneys from reviewing discovery, discovery reforms would fail to meet their goals, and defendants would receive sub-standard representation. The recent adoption of digital evidence platforms by local jurisdictions allows us to study whether defense attorneys consistently fulfill their duty to review discovery. Analyzing data from digital evidence platforms used in felony cases in four Texas counties between 2018 and 2020, we examine whether and when defense attorneys fail to access evidence disclosed by the prosecution. We find that a substantial number of defense attorneys never access the discovery. The access rate varies by county, offense seriousness, attorney category, attorney experience, and file type. Drawing on review of prior scholarship and Bayesian analysis of the data, we discuss plausible interpretations of these variations
Context-Switch Attacks: Understanding and Mitigating the Threat to LLM Applications
No full textLarge Language Models (LLMs) are transforming conversational AI, yet their dependence on prompt-supplied context exposes them to context-switch attacks that covertly steer dialogue toward sensitive or malicious ends. A 70 one-sided conversation transcript evaluation set was constructed spanning various fraudulent scenarios. Each transcript embeds adversarial patterns drawn while preserving natural conversational flow. We introduce a hybrid defense that pairs a BERT-based semantic-drift detector (cosine-similarity threshold = 0.70) with a curated keyword and hack-phrase scanner to counter these threats. In aggregate, the system delivered 100 % recall, intercepting every simulated phishing or data-harvesting attempt. The keyword layer achieved perfect precision, generating a mean of 1.93 alerts per transcript with zero false positives. In contrast, the semantic layer contributed a mean of 1.84 additional warnings and captured all four attacks that lacked sensitive keywords. Overall, conversations triggered 7.5 risk signals on average (≈ 1.1 per message), and 98.6 % of transcripts activated at least two independent alarms, evidencing robust redundancy. The principal trade-off surfaced in the semantic component, where roughly one-third of its warnings reflected benign pivots, such as address or insurance confirmations, highlighting the tension between maximal coverage and conversational fluidity. Building on these findings, we recommend adaptive similarity thresholds, multistage escalation, and user-configurable sensitivity profiles to balance security and usability. By documenting the mechanics and impact of context-switch attacks and demonstrating an adequate dual-layer safeguard, this work provides both an empirical foundation and practical guidance for hardening LLM-based systems deployed in high-stakes, real-world environments
Enhancing News Article Generation with AI Tools
No full textAbstract. This paper aims to present a comparative analysis of a custom-built AI Journalist Assistant and ChatGPT 4.0 for news article generation. The goal is to evaluate the performance of each model based on accuracy, speed, ethical safeguards, and relevance, particularly in the context of journalism. While ChatGPT is widely used for general-purpose content creation, its reliance on older data and potential for plagiarism presents challenges in the fast-paced, high-stakes world of news reporting. To address these issues, we will design and implement an AI Assistant using Retrieval-Augmented Generation (RAG) techniques, focusing on real-time data access, bias reduction, and plagiarism prevention. By comparing both models through established metrics such as BERTScore, precision, recall, and inference time, this paper contributes insights into the role of specialized AI tools in journalistic workflows and highlights the trade-offs between prompt engineering and domain-specific AI development. The findings will guide future advancements in AI-generated news and its ethical implications
Airplane!: Does the Federal Aviation Act Preempt State Law Design Defect Claims?
No full textWhich level of government—state or federal—has jurisdiction to set safety standards for aviation defect design is a nuanced issue. Generally speaking, federal law preempts state law in this context by occupying the field through a regulatory scheme that permeates aviation safety. The Federal Aviation Act of 1958 (FAAct) confers its namesake agency—the Federal Aviation Administration (FAA)—the powers to prescribe the “minimum standards required in the interest of safety.” The word “minimum” has been seized upon by those on the state side of the debate to advocate that it means the floor. And they contend that those in the Union have the freedom to set so-called higher standards—the ceiling. Granted, this may be a persuasive argument in some cases, but not here. First, the FAA has a five-stage type certification process. Further, there is a production certificate and airworthiness certificate. Second, the United States has executed treaties with foreign nations over reciprocity regarding aircraft safety standards.
Airplane accident cases are emotionally charged due to their propensity to involve fatalities—the argumentum ad passiones (appeal to emotion) fallacy. This emotional response is papering over underlying issues. If an airplane has obtained certification and tragedy strikes, there may be other avenues to explore. Negligence and tort claims come to mind. Preemption of design defect claims is limited to that context. A broader holding could sweep up remedies at law that should remain available.
As of the writing of this Article, impossibility preemption is premature but looms large. If all fifty states are permitted to enact aviation safety design laws, compliance would become problematic because air travel transcends state boundaries. Additionally, conflict preemption could arise. Divergence in design defects may place plaintiffs in the position of “heads I win, tails you lose.” Adhering to the standards for one state may fall below the threshold for another, or their frameworks may be at odds. Not to mention federal requirements
Why Domestic Violence Offenders Don\u27t Give Up Their Guns
No full textPerpetrators of intimate partner violence are barred by federal law and many states\u27 laws from possessing firearms. While such prohibitions enjoy popular support, they are sporadically and inconsistently enforced, placing the lives of survivors at risk when offenders do not voluntarily comply. Many experts, including this author, have offered legal and policy solutions to increase the likelihood that perpetrators of domestic violence will either willingly relinquish their guns or otherwise be dispossessed of them. But these proposals may have been premature.
This Article is the first to take a step back and inquire why offenders do not surrender their firearms as ordered and what might incentivize them to do so. Understanding the worldviews and lived experiences of those subject to domestic violence gun prohibitions is a critical, and to-date ignored, first step to formulating viable solutions.
The Article is based on original empirical research conducted with perpetrators of intimate partner violence enrolled in a Batterer Intervention Program in Texas. The research reveals a range of novel findings based on both survey responses and in-depth interviews. The data show that most men subject to domestic violence firearm regulations are aware that they are prohibited possessors. They are reluctant to comply with the law, however, due to the atypically high levels of violence-including gun violence-they have experienced in their lifetimes, which has led them to conclude that firearms are necessary to protect themselves and their loved ones from harm. Respondents also identify closely with a stereotypically masculine identity that leads them to associate gun ownership with power and control; have strong (but not necessarily accurate) opinions about the Second Amendment; and are enmeshed in cultures where gun-carrying is the norm.
These insights about the men who are impacted by domestic violence gun regulations can help us promulgate laws and policies that offenders will be more inclined to comply with and are more likely to be enforced and enforceable. And importantly, successfully removing firearms from the hands of abusers can bring us one step closer to ending the epidemic of intimate partner violence gun fatalities in the United States
Enhancing Inquiry-Based Science Instruction: The Role of Professional Learning Communities and Instructional Coaching for Elementary Science Teachers
No full textInquiry-based science instruction fosters critical thinking, problem-solving, and scientific literacy by engaging students in exploration, questioning, and reflective learning. However, implementing inquiry-based practices presents significant challenges for elementary teachers, who often balance multiple subjects and may feel less confident facilitating open-ended investigations. This study examines how school-based coaching and professional learning communities (PLCs) support teachers in adopting inquiry-based science instruction. Grounded in distributed cognition theory of learning, this research draws on qualitative data from classroom observations and artifacts from coaching sessions and PLC meetings. Using a multiple case study design, this study identifies four case studies to illustrate varying levels of alignment between PLC discussions and classroom implementation. Findings emphasize the importance of sustained, flexible professional development and coaching that integrates collaborative learning into consistent classroom practice. Distributing cognitive demands through coaching and PLCs strengthens inquiry-based instruction and enhances science teaching in diverse educational contexts. This study offers insights for educators, school leaders, and policymakers aiming to advance student-centered science education
Econometric Analysis of Social Interactions and Decision-Making
No full textThis dissertation explores the econometric analysis of social interactions and decision-making. The first two chapters advance econometric methodology by addressing key challenges in identifying and estimating peer effects, specifically tackling issues related to partial network data and endogenous link formation. The third chapter presents an empirical analysis of parental over-aspiration and its impact on children\u27s academic outcomes, using monotone instrumental variables constructed from social network structures.
Chapter 1 is Identification and Estimation of Discrete Choice Models with Spillovers Using Partial Network Data . This paper investigates peer effects in discrete choice models with incomplete data on social links. We consider an undirected dyadic link formation model in which connections arise from homophily—similarities in individual characteristics—and unobserved individual fixed effects. Homophily effects are identified using the observed patterns of link formation among tetrads (groups of four individuals), while the distribution of individual fixed effects is identified through the configurations observed among triads (groups of three individuals). I propose an estimation strategy based on the estimated link formation probabilities to study peer effects on individual decision-making and establish its large sample properties. Simulations illustrate that the finite sample performance of the estimator is close to that obtained when the true network were known. Using data on household microfinance participation in rural India from Banerjee et al. (2013), where network data are available, I detect positive peer effects even with partial network data.
Chapter 2 is Inference for Social Interactions in Large Endogenous Networks (co-authored with Wan Zhang). We study the identification and estimation of social interactions in large endogenous networks. Our analysis focuses on binary-action games of incomplete information in which an agent\u27s expected payoff relies on her characteristics, peers\u27 average characteristics, the average of her beliefs about peers\u27 actions, and some preference shocks. Endogeneity in networks results from unobserved characteristics that affect both link formation and individual decision-making. To identify the utility functions, we express the unobserved characteristics as some unknown function of observed variables and address the endogeneity issue through a control function approach. The identification strategy holds even if multiple equilibria exist. We employ the strategy to develop a semiparametric estimator and assess finite-sample performance through Monte Carlo simulations.
Chapter 3 is Echoes of Icarus: How Parental Over-Aspiration Shapes Children\u27s Academic Success . This paper documents the negative influence of overly ambitious parental aspirations on children\u27s academic outcomes, utilizing data from the Education Longitudinal Study 2002. After controlling for family socioeconomic status, I find that parents with better education are less likely to form overly ambitious aspirations, which represents a new channel to account for the positive effects of parental education on children\u27s academic performances. To address the possibility that parental aspirations are endogenous, I construct instruments based on children\u27s social network. For robustness, I use an alternative approach based on bounding the effect of parental aspirations under weaker assumptions. The results are consistent across both approaches
Exploring The Accreditation Lever: A Multi-Method Examination Of How Accreditation Policy Relates To College Outcomes And Transfer Student Baccalaureate Attainment
No full textThis dissertation is a multi-method approach with a 3-paper format including three related but distinct studies. The first paper utilizes qualitative document analysis. Findings illuminated differences in how accreditors address transfer outcomes and transfer credit, signaling important inconsistencies across accreditors and inconsistencies in accreditation requirements for HEOs that answer to different accreditors. The second paper utilizes descriptive statistics and linear regression models to explore the relationship between HEO transfer graduation rates and accreditors and their policies. Findings illuminate statistically significantly differences in transfer graduation rates between accreditors suggesting that accreditation does matter for transfer outcomes. Finally, the third paper focuses on a continuous improvement accreditation requirement, the Quality Enhancement Plan (QEP). Using document analysis and interviews with QEP leaders at community colleges, findings illuminate how the QEP works as a catalyst for organizational learning that leads to outcomes. The QEP can be focused on any student outcome, including transfer outcomes if a HEO chooses, making this finding particularly compelling insight for how accreditation can improve student outcomes at HEOs. Collectively, the findings of this dissertation illuminate previously unknown relationship between accreditation and transfer outcomes and provide a methodological map for further research on accreditation policy, an understudied area of policy